Updated on 2024/12/21

写真a

 
MATSUMOTO, Takashi
 
Affiliation
Faculty of Science and Engineering
Job title
Professor Emeritus
Degree
Doctor of Engineering ( Waseda University )
MSc ( Harvard University )

Research Experience

  • 2003.04
    -
    2004.03

    Cambridge University   Department of Engineering   Visiting Scientist

  • 1977
    -
    1979

    米国カリフォルニア大学バークレー電気工学・計算機科学 客員研究員

  • 1977
    -
    1979

    Visting Scientist, Department of Electrical Engineering and Computer Sciences,

  •  
     
     

    University of California, Berkeley

Education Background

  •  
    -
    1966

    Waseda University  

  •  
    -
    1966

    Waseda University   Faculty of Engineering  

Committee Memberships

  • 1995
    -
     

    映像メディア学会  次世代画像入力研究会委員

  • 1993
    -
     

    電気学会  カオス調査専門委員会委員長

  • 1992
    -
     

    Circuits,Systems and Signal Processing  Associate Editor

  • 1992
    -
     

    IEEE CAS Society  Board of Governors

  • 1990
    -
    1991

    電子通信学会  誌論文委員,誌編集委員会委員,非線形問題専門委員会専門委員長

  • 1990
    -
     

    電子情報通信学会  非線形問題研究専門委員会委員長

  • 1988
    -
    1989

    IEEE CAS Society  Associate Editor,Guest Co-editor

  • 1981
    -
    1983

    IEEE CAS Socity  Associate Editor,Guest Co-editor

  • 1976
    -
    1977

    計測自動制御学会  誌編集委員,論文集委員

▼display all

Professional Memberships

  •  
     
     

    映像情報メディア学会

  •  
     
     

    情報処理学会

  •  
     
     

    ヒューマンインターフェイス学会

  •  
     
     

    映像メディア学会

  •  
     
     

    Systems and Signal Processing

  •  
     
     

    Circuits

  •  
     
     

    IEEE CAS Society

  •  
     
     

    IEEE CAS Socity

  •  
     
     

    電子通信学会

  •  
     
     

    計測自動制御学会

  •  
     
     

    電気学会

  •  
     
     

    日本神経回路学会

  •  
     
     

    電子情報通信学会

  •  
     
     

    IEEE

  •  
     
     

    Institute of Image Information and Television Engineers

  •  
     
     

    Information Processing Society of Japan

  •  
     
     

    Human Interface Society

  •  
     
     

    The Society of Instrument and Control Engineers

  •  
     
     

    Japanese Neural Network Society

  •  
     
     

    Information and Communication Engineers

  •  
     
     

    Institute of Electronics

  •  
     
     

    IEEE

▼display all

Research Areas

  • Control and system engineering / Communication and network engineering / Life, health and medical informatics / Intelligent robotics / Perceptual information processing

Research Interests

  • Signature Verification

  • 複雑系

  • 逐次モンテカルロ

  • マルコフ連鎖モンテカルロ

  • 情報処理

  • 信号処理

  • Bayesian Learning

  • complex systems

  • Sequential Monte Carlo

  • Markov Chain Monte Carlo

  • Signal Processing

▼display all

Awards

  • Life Fellow

    2010   IEEE  

    Winner: MATSUMOTO Takashi

  • IEEE Life Fellow

    2010  

  • The Second IP Design Award Development

    2000  

    Winner: MATSUMOTO Takashi

  • IEEE Certificate of Appreciation

    1995   IEEE CAS Society  

    Winner: MATSUMOTO Takashi

  • Best Paper Award

    1994   Japanese Society of Neural Networks  

    Winner: MATSUMOTO Takashi

  • 日本神経回路学会 論文賞

    1994  

  • Fellow

    1984   IEEE  

    Winner: MATSUMOTO Takashi

  • IEEE Fellow

    1984  

▼display all

 

Books and Other Publications

  • EE Text 基礎電気回路

    オーム社  2011 ISBN: 9784274211232

  • 階層ベイズモデルとその周辺ー統計科学のフロンティア4(第3刷)

    岩波書店  2006 ISBN: 400006844X

  • 複雑系叢書(全7巻)1.複雑系の構造と予測

    共立出版  2006 ISBN: 4320034457

  • Bifurcations: Sights, Sounds and Mathematics

    Springer - Verlag  1993

  • Bifurcations: Sights, Sounds and Mathematics

    Springer - Verlag  1993

  • カオス"第2章"カオスを電子回路で捕える

    サイエンス社刊  1990

▼display all

Research Projects

  • NIRSデータによる脳信号解読

    Project Year :

    2010
    -
     
     

     View Summary

    前頭前野酸化ヘモグロビン、脱酸化ヘモグロビン量から人の不安度を予測する

  • Hyperspectral Imagingによる生体認証

    Project Year :

    2010
    -
     
     

     View Summary

    Hyperspectral Imagingにより個人認証を行う

  • NIRS-based Brain Sgnal Decoding

    Project Year :

    2010
    -
     
     

     View Summary

    Predict anxiety index from NIRS data

  • Hyperspectral Imaging-based Biometric Authentication

    Project Year :

    2010
    -
     
     

     View Summary

    Authenticate individuals by hyperspectral imaging

  • ノンパラメトリック Bayes学習

    Project Year :

    2009
    -
     
     

     View Summary

    Stick-Breaking PriorをもつBayes学習とその実装手段としてのGEM

  • 脳波信号解読

    Project Year :

    2009
    -
     
     

     View Summary

    脳波から人の意図を読み取る

  • Nonparameteric Bayesian Learning

    Project Year :

    2009
    -
     
     

     View Summary

    Bayesian approach with Stick-Breaking GEM implementation

  • EEG-based Brain Signal Decoding

    Project Year :

    2009
    -
     
     

     View Summary

    prediction of intention from EEG

  • Dynamic Gaussian Process 遺伝子発現ネットワーク予測

    Project Year :

    2005
    -
     
     

     View Summary

    Dynamic Gaussian Processにより遺伝子発現ネットワークを予測する

  • Gene expression network prediction by Dynamic Gaussian Processes

    Project Year :

    2005
    -
     
     

     View Summary

    Predict gene regulatory network by Dynamic Gaussian Processes

  • アナログ回路設計の研究

    Project Year :

    2000
    -
     
     

  • 複雑系の解明と新技術の開発

    Project Year :

    1999
    -
     
     

  • 複雑系の解明と新技術の開発

    受託研究

    Project Year :

    1998
    -
     
     

  • モンテカルロ HMM スポーツゲーム活況度予測

     View Summary

    スポーツビデオ動画像からゲームの活況度を予測する

  • モンテカルロHMM「膜タンパク質構造予測」

  • Monte Carlo HMM Sports Game Activity Prediction

     View Summary

    predict activity associated with sports video

  • Monte Carlo HMM, Membrane Protein Structure Prediction

▼display all

Misc

  • ベイズ逐次学習によるBrain-Machine Interfaceのユーザビリティ向上

    重住周, 芹沢央子, 松本隆

    電子情報通信学会論文誌   J97-D ( 1 ) 62 - 74  2014

  • Expectation-maximization algorithms for inference in Dirichlet processes mixture

    T. Kimura, T. Tokuda, Y. Nakada, T. Nokajima, T. Matsumoto, A. Doucet

    PATTERN ANALYSIS AND APPLICATIONS   16 ( 1 ) 55 - 67  2013.02

     View Summary

    Mixture models are ubiquitous in applied science. In many real-world applications, the number of mixture components needs to be estimated from the data. A popular approach consists of using information criteria to perform model selection. Another approach which has become very popular over the past few years consists of using Dirichlet processes mixture (DPM) models. Both approaches are computationally intensive. The use of information criteria requires computing the maximum likelihood parameter estimates for each candidate model whereas DPM are usually trained using Markov chain Monte Carlo (MCMC) or variational Bayes (VB) methods. We propose here original batch and recursive expectation-maximization algorithms to estimate the parameters of DPM. The performance of our algorithms is demonstrated on several applications including image segmentation and image classification tasks. Our algorithms are computationally much more efficient than MCMC and VB and outperform VB on an example.

    DOI

  • 4-hydroxy-2-nonenal によるTDP-43の細胞内局在,凝集体形成,リン酸化の変化

    向野佳奈子, 畑中悠佑, 松本隆, 和田圭司, 株田智弘

    日本分子生物学会第36回年会    2013

  • Domain-Dependent/Independent Topic Switching Model for Online Reviews with Numerical Ratings

    Yasutoshi Ida, Takuma Nakamura, Takashi Matsumot

    CIKM 2013     229 - 238  2013

    DOI

  • A comparative study of ASSR classification problem using bipolar and monopolar EEG voltages

    Fumi Fukaya, Takashi Nakamura, Hironao Namba, Takashi Matsumoto

    2013 International Conference on Brain and Health Informatics    2013

  • Label-Related/Unrelated Topic Switching Model: A Partially Labeled Topic Model Handling Infinite Label-unrelated Topics

    Yasutoshi Ida, Takuma Nakamura, Takashi Matsumoto

    ACPR2013(RACVPR2013)    2013

  • 構文構造を考慮した特徴量を用いたトピックモデルによる評判分析

    横井創磨, 井田安俊, 小笠原光貴, 松本隆

    第16回情報論的学習理論ワークショップ (IBIS2013)    2013

  • 署名データ補正とスコア統合によるカメラに基づくオンライン署名 認証の高精度化

    村松大吾, 安田久美子, 松本隆, 八木康史

    電子情報通信学会論文誌A   J96-A ( 12 ) 780 - 789  2013

  • Domain-Dependent/Independent Topic Switching Model for Online Reviews with Numerical Ratings

    Yasutoshi Ida, Takuma Nakamura, Takashi Matsumot

    CIKM 2013     229 - 238  2013

    DOI

  • Classification of Auditory Steady-State Responses to Speech Data

    T. Nakamura, H. Namba, T. Matsumoto

    2013 6TH INTERNATIONAL IEEE/EMBS CONFERENCE ON NEURAL ENGINEERING (NER)     1025 - 1028  2013

     View Summary

    This paper presents an auditory steady-state response (ASSR)-based brain-computer interface (BCI) that uses artificial speech data synthesized by a text-to-speech (TTS) system. Many ASSR-based BCI systems that use pure tone (sinusoid) or an abrupt beep as auditory stimuli have been proposed. However, while these systems have achieved high classification accuracy, our group has found that participants find the monotonous stimuli to be hypnotic and annoying. Practical BCI systems should have user-friendly designs. Thus, as a first step, we develop a new experimental BCI system in which we change the type of stimuli from pure tone carrier to artificial speech data, which are clear enough for participants to recognize the meaning of sentences. With eight participants, the average accuracy of the system is 78.6 +/- 5.32% for the binary classification problem. This suggests that the proposed system can be used in practical BCI.

    DOI

  • A comparative study of ASSR classification problem using bipolar and monopolar EEG voltages

    Fumi Fukaya, Takashi Nakamura, Hironao Namba, Takashi Matsumoto

    2013 International Conference on Brain and Health Informatics    2013

  • Label-Related/Unrelated Topic Switching Model: A Partially Labeled Topic Model Handling Infinite Label-unrelated Topics

    Yasutoshi Ida, Takuma Nakamura, Takashi Matsumoto

    ACPR2013(RACVPR2013)    2013

  • Monitoring of brain function by near infrared spectroscopy (NIRS) for prevention of brain diseases

    Kaoru Sakatani, Naohiro Takemura, Wakana Ishikawa, Yukikatsu Fukuda, Takashi Matsumoto

    35th Annual Int'l Conference of the IEEE Engineering in Medicine and Biology Society    2013

  • Classification of Auditory Steady-State Responses Incorporating Alpha Waves

    H. Namba, T. Nakamura, T. Matsumoto

    5th International BCI Meeting    2013

    DOI

  • A Hidden Markov Model Approach for Onset Detection in Music Signals

    高松愛, 小泉幸広, 松本隆

    情報処理学会第75回全国大会   ( 2 ) 509 - 510  2013

  • 0.93nm波長解像度ハイパースペクトラルDatacube生体認証

    佐藤優太, 赤澤史嗣, 村松大吾, 松本隆, 中村厚, 宗田孝之

    電子情報通信学会 総合大会 AS-4-1     S38-S39  2013

  • Bayesian Sequential Learning for EEG-based BCI Classification Problems

    S.Shigezumi, H.Hara, H.Namba, C.Serizawa, Y.Dobashi, A.Takemoto, K.Nakamura, T.Matsumoto

    Brain-Computer Interface Systems - Recent Progress and Future Prospects (Chapter 4)     61 - 89  2013

    DOI

  • Online Bayesian Learning with Natural Sequential Prior Distribution

    Y. Nakada, M. Wakahara, T. Matsumoto

    IEEE Transactions on Neural Networks and Learning Systems     1 - 15  2013

    DOI

  • Bayesian prediction of anxiety level in aged people at rest using 2-channel NIRS data from prefrontal cortex

    Yukikatsu Fukuda, Wakana Ishikawa, Ryuhei Kanayama, Takashi Matsumoto, Naohiro Takemura, Kaoru Sakatani

    The 41st Annual ISOTT Meeting    2013

    DOI

  • Improving Information-Transfer Rate of Auditory Steady-State Responses Using Alpha Waves

    T. Nakamura, H. Namba, T. Matsumot

    35th Annual International Conference of the IEEE Engineering in Medicine and Biology Society    2013

  • Monitoring of brain function by near infrared spectroscopy (NIRS) for prevention of brain diseases

    Kaoru Sakatani, Naohiro Takemura, Wakana Ishikawa, Yukikatsu Fukuda, Takashi Matsumoto

    35th Annual Int'l Conference of the IEEE Engineering in Medicine and Biology Society    2013

  • Classification of Auditory Steady-State Responses Incorporating Alpha Waves

    H. Namba, T. Nakamura, T. Matsumoto

    5th International BCI Meeting    2013

    DOI

  • Bayesian Sequential Learning for EEG-based BCI Classification Problems

    S.Shigezumi, H.Hara, H.Namba, C.Serizawa, Y.Dobashi, A.Takemoto, K.Nakamura, T.Matsumoto

    Brain-Computer Interface Systems - Recent Progress and Future Prospects (Chapter 4)     61 - 89  2013

    DOI

  • An Authentication Method by High Spectral Resolution Palm Datacube

    Yuta Sato, Fumitsugu Akazawa, Daigo Muramatsu, Takashi Matsumoto, Atsushi Nakamura, Takayuki Sota

    2013 INTERNATIONAL CONFERENCE ON BIOMETRICS AND KANSEI ENGINEERING (ICBAKE)     239 - 244  2013

     View Summary

    A biometric authentication method is proposed based on hyperspectral image data derived from the palm of the hand. The data are acquired using a recently developed device that captures reflectance across the 396.37-990.64 nm range with a spectral resolution of 0.93 nm. The acquired image data represent the distributions of various biological substances. First, the proposed method computes the spatial correlations between the test data and a set of stored template data for each wavelength of the spectrally resolved image. Next, the algorithm computes a score by integrating the spatial correlations for each wavelength. This method incorporates the well-known vein structure in the NIR range and other rather intricate structures located in the shallower parts of hands. In the evaluation, 1500 data were acquired from 30 subjects and the equal error rate was 0.611%.

    DOI

  • Online Bayesian Learning with Natural Sequential Prior Distribution

    Y. Nakada, M. Wakahara, T. Matsumoto

    IEEE Transactions on Neural Networks and Learning Systems     1 - 15  2013

    DOI

  • Bayesian prediction of anxiety level in aged people at rest using 2-channel NIRS data from prefrontal cortex

    Yukikatsu Fukuda, Wakana Ishikawa, Ryuhei Kanayama, Takashi Matsumoto, Naohiro Takemura, Kaoru Sakatani

    The 41st Annual ISOTT Meeting    2013

    DOI

  • Improving Information-Transfer Rate of Auditory Steady-State Responses Using Alpha Waves

    T. Nakamura, H. Namba, T. Matsumot

    35th Annual International Conference of the IEEE Engineering in Medicine and Biology Society    2013

  • Domain-dependent/independent topic switching model for online reviews with numerical ratings

    Yasutoshi Ida, Takuma Nakamura, Takashi Matsumoto

    International Conference on Information and Knowledge Management, Proceedings     229 - 238  2013

     View Summary

    We propose a domain-dependent/independent topic switching model based on Bayesian probabilistic modeling for modeling online product reviews that are accompanied with numerical ratings provided by users. In this model, each word is allocated to a domain-dependent topic or a domain-independent topic, and the distribution of topics in an online review is connected to an observed numerical rating via a linear regression model. Domain-dependent topics utilize domain information observed with a corpus, and domain-independent topics utilize the framework of Bayesian Non-parametrics, which can estimate the number of topics in posterior distributions. The posterior distribution is estimated via collapsed Gibbs sampling. Using real data, our proposed model had smaller mean square error and smaller average mean error with a small model size and achieved convergence in fewer iterations for a regression task involving online review ratings, outperforming a baseline model that did not consider domains. Moreover, the proposed model can also tell us whether the words are positive or negative in the form of continuous values. This feature allows us to extract domain-dependent and -independent sentiment words. Copyright is held by the owner/author(s).

    DOI

  • Bayesian STAI Anxiety Index Predictions Based On Prefrontal Cortex NIRS Data for the Resting State

    Masakaze Sato, Wakana Ishikawa, Tomohiko Suzuki, Takashi Matsumoto, Takeo Tsujii, Kaoru Sakatani

    Advances in Experimental Medicine and Biology, Springer   765   251 - 256  2013

    DOI

  • Classification of Auditory Steady-State Responses Incorporating Alpha Waves

    H. Namba, T. Nakamura, T. Matsumoto

    5th International BCI Meeting   accepted  2013

    DOI

  • 0.93nm波長解像度ハイパースペクトラルDatacube生体認証

    佐藤優太, 赤澤史嗣, 村松大吾, 松本隆, 中村厚, 宗田孝之

    電子情報通信学会 総合大会 AS-4-1     S  2013

  • 前頭前野2チャンネルNIRS時系列データの定常化によるSTAI不安度予測

    金山龍平, 石川わかな, 福田行克, 山本篤, 細野晴実, 竹村尚大, 酒谷薫, 松本隆

    電子情報通信学会 総合大会 「学生ポスターセッション」*優秀ポスター賞    2013

  • 前頭前野安静時NIRSデータによるSTAI不安度予測の試み

    細野晴実, 石川わかな, 福田行克, 金山龍平, 山本篤, 酒谷薫, 武村尚大, 松本隆

    電子情報通信学会総合大会「学生ポスターセッション」    2013

  • Bayesian Sequential Learning for EEG-based BCI Classification Problems

    S.Shigezumi, H.Hara, H.Namba, C.Serizawa, Y.Dobashi, A.Takemoto, K.Nakamura, T.Matsumoto

    Brain-Computer Interface Systems - Recent Progress and Future Prospects (Chapter 4)    2013

    DOI

  • Bayesian STAI Anxiety Index Predictions Based On Prefrontal Cortex NIRS Data for the Resting State

    Masakaze Sato, Wakana Ishikawa, Tomohiko Suzuki, Takashi Matsumoto, Takeo Tsujii, Kaoru Sakatani

    Advances in Experimental Medicine and Biology, Springer   765   251 - 256  2013

    DOI

  • Nonparametric bayes-based heterogeneous "Drosophila Melanogaster" gene regulatory network inference: T-process regression

    Hiroki Miyashita, Takuma Nakamura, Yasutoshi Ida, Takashi Matsumoto, Takashi Kaburagi

    IASTED Multiconferences - Proceedings of the IASTED International Conference on Artificial Intelligence and Applications, AIA 2013     51 - 58  2013

     View Summary

    Recent research into time-varying network inference for gene expression data mainly assumes that gene regulatory networks have linear interactions. This assumption is straightforward and requires comparatively simple model building. However, in various previous biological studies, gene expression data have been believed to have nonlinear properties in their regulatory interactions. To address this, we adopted a nonparametric Bayesian regression method (e.g. a Gaussian Process) for predicting interactions into a time-varying network to achieve more flexible regression capability. The proposed method was evaluated on Drosophila melanogaster gene data, which has been used as a benchmark in a number of studies. This dataset, which was measured by a microarray test, is known to include noise. To obtain stronger robustness to noisy data, in our algorithm, we employed the T-Process instead of the conventional Gaussian Process. To the best of our knowledge, this is the first algorithm to apply nonparametric Bayesian regression method to a time-varying gene regulatory network problem. Our basic algorithm employed reversible jump Markov Chain Monte Carlo (RJMCMC) for inference of whole network structures. The method can handle the two inference problems: (i) change point detection and (ii) network structure inference simultaneously.

    DOI

  • Classification of Auditory Steady-State Responses Incorporating Alpha Waves

    H. Namba, T. Nakamura, T. Matsumoto

    5th International BCI Meeting   accepted  2013

    DOI

  • Bayesian Sequential Learning for EEG-based BCI Classification Problems

    S.Shigezumi, H.Hara, H.Namba, C.Serizawa, Y.Dobashi, A.Takemoto, K.Nakamura, T.Matsumoto

    Brain-Computer Interface Systems - Recent Progress and Future Prospects (Chapter 4)    2013

    DOI

  • An Authentication Method by High Spectral Resolution Palm Datacube

    Yuta Sato, Fumitsugu Akazawa, Daigo Muramatsu, Takashi Matsumoto, Atsushi Nakamura, Takayuki Sota

    2013 INTERNATIONAL CONFERENCE ON BIOMETRICS AND KANSEI ENGINEERING (ICBAKE)   accepted   239 - 244  2013

     View Summary

    A biometric authentication method is proposed based on hyperspectral image data derived from the palm of the hand. The data are acquired using a recently developed device that captures reflectance across the 396.37-990.64 nm range with a spectral resolution of 0.93 nm. The acquired image data represent the distributions of various biological substances. First, the proposed method computes the spatial correlations between the test data and a set of stored template data for each wavelength of the spectrally resolved image. Next, the algorithm computes a score by integrating the spatial correlations for each wavelength. This method incorporates the well-known vein structure in the NIR range and other rather intricate structures located in the shallower parts of hands. In the evaluation, 1500 data were acquired from 30 subjects and the equal error rate was 0.611%.

    DOI

  • Nonparametric bayes-based heterogeneous "Drosophila Melanogaster" gene regulatory network inference: T-process regression

    Hiroki Miyashita, Takuma Nakamura, Yasutoshi Ida, Takashi Matsumoto, Takashi Kaburagi

    IASTED Multiconferences - Proceedings of the IASTED International Conference on Artificial Intelligence and Applications, AIA 2013     51 - 58  2013

     View Summary

    Recent research into time-varying network inference for gene expression data mainly assumes that gene regulatory networks have linear interactions. This assumption is straightforward and requires comparatively simple model building. However, in various previous biological studies, gene expression data have been believed to have nonlinear properties in their regulatory interactions. To address this, we adopted a nonparametric Bayesian regression method (e.g. a Gaussian Process) for predicting interactions into a time-varying network to achieve more flexible regression capability. The proposed method was evaluated on Drosophila melanogaster gene data, which has been used as a benchmark in a number of studies. This dataset, which was measured by a microarray test, is known to include noise. To obtain stronger robustness to noisy data, in our algorithm, we employed the T-Process instead of the conventional Gaussian Process. To the best of our knowledge, this is the first algorithm to apply nonparametric Bayesian regression method to a time-varying gene regulatory network problem. Our basic algorithm employed reversible jump Markov Chain Monte Carlo (RJMCMC) for inference of whole network structures. The method can handle the two inference problems: (i) change point detection and (ii) network structure inference simultaneously.

    DOI

  • PreGO: A Protein Function Prediction Algorithm Based on an Infinite Mixture of Hidden Markov and Bayesian Network Models

    T. Kaburagi, Y. Koizumi, K. Oota, T. Matsumoto

    BiCOB-2013    2013

  • Classification of Auditory Steady-State Responses Incorporating Alpha Waves

    H. Nanba, T. Nakamura, T. Matsumoto

    5th International BCI Meeting   accepted  2013

    DOI

  • A Hidden Markov Model Approach for Onset Detection in Music Signals

    高松愛, 小泉幸広, 松本隆

    情報処理学会第75回全国大会     2  2013

  • 画像領域分割問題に対する階層ディリクレ過程事前分布マルコフ確率場の提案

    岸悠介, 中村拓磨, 原田竜弘, 松本隆

    電子情報通信学会 信学技報IBISML2012-105   112 ( 454 ) 87 - 94  2013

  • アルファ波成分を用いた脳波ASSRデータの2クラス判別

    南波寛直, 中村尭, 松本隆

    電子情報通信学会 信学技報MBE2012-104   112 ( 479 ) 79 - 82  2013

  • 0.93nm波長解像度ハイパースペクトラルDatacube生体認証

    佐藤優太, 赤澤史嗣, 村松大吾, 松本隆, 中村厚, 宗田孝之

    電子情報通信学会 総合大会 AS-4-1    2013

  • 前頭前野2チャンネルNIRS時系列データの定常化によるSTAI不安度予測

    金山龍平, 石川わかな, 福田行克, 山本篤, 細野晴実, 竹村尚大, 酒谷薫, 松本隆

    電子情報通信学会 総合大会 「学生ポスターセッション」    2013

  • DFT・CCA を用いた脳波 SSVEP 2 クラス判別問題における時間窓の検討

    谷部嘉純, 重住周, 松本隆

    電子情報通信学会 総合大会「学生ポスターセッション」    2013

  • Nonparametric Bayes-based Heterogeneous “Drosophila melanogaster” Gene Regulatory Network Inference: T-Process Regression

    H. Miyashita, T. Suzuki, T. Nakamura, Y. Ida, T. Matsumoto, T. Kaburagi

    AIA2013    2013

    DOI

  • PreGO: A Protein Function Prediction Algorithm Based on an Infinite Mixture of Hidden Markov and Bayesian Network Models

    T. Kaburagi, Y. Koizumi, K. Oota, T. Matsumoto

    BiCOB-2013    2013

  • Classification of Auditory Steady-State Responses Incorporating Alpha Waves

    H. Nanba, T. Nakamura, T. Matsumoto

    5th International BCI Meeting   accepted  2013

    DOI

  • Scene detection using a large number of text features

    Ichiro Yamada, Yohei Nakada, Atsushi Matsui, Takashi Matsumoto, Kikuka Miura, Hideki Sumiyoshi, Masahiro Shibata, Nobuyuki Yagi

    ITE Transactions on Media Technology and Applications   1 ( 2 ) 157 - 166  2013

     View Summary

    Broadcasting stations store a large volume of TV programs and manage them in their archives. To enable such programs to be used effectively, the technique for analyzing what is depicted in each scene plays a crucial role. TV programs often contain typical scenes which are used for specific purposes. This paper proposes a novel method of detecting such typical scenes by analyzing the context of closed captions. The proposed method handles a huge number of text features extracted from the closed captions through its use of a Monte Carlo based boosting algorithm. In experiments, we classified text segments extracted from the closed captions as to whether or not the corresponding scene is typical one. The results confirmed that our method classified with comparable accuracy to a conventional method using the AdaBoost algorithm and achieved a dramatic reduction in the learning time.

    DOI

  • Monitoring of brain function by near infrared spectroscopy (NIRS) for prevention of brain diseases

    Kaoru Sakatani, Naohiro Takemura, Wakana Ishikawa, Yukikatsu Fukuda, Takashi Matsumoto

    35th Annual Int'l Conference of the IEEE Engineering in Medicine and Biology Society   accepted  2013

  • PreGO: A Protein Function Prediction Algorithm Based on an Infinite Mixture of Hidden Markov and Bayesian Network Models

    T. Kaburagi, Y. Koizumi, K. Oota, T. Matsumoto

    BiCOB-2013   accepted  2013

  • Monitoring of brain function by near infrared spectroscopy (NIRS) for prevention of brain diseases

    Kaoru Sakatani, Naohiro Takemura, Wakana Ishikawa, Yukikatsu Fukuda, Takashi Matsumoto

    35th Annual Int'l Conference of the IEEE Engineering in Medicine and Biology Society   accepted  2013

  • PreGO: A Protein Function Prediction Algorithm Based on an Infinite Mixture of Hidden Markov and Bayesian Network Models

    T. Kaburagi, Y. Koizumi, K. Oota, T. Matsumoto

    BiCOB-2013   accepted  2013

  • Nonparametric bayes-based heterogeneous "Drosophila Melanogaster" gene regulatory network inference: T-process regression

    Hiroki Miyashita, Takuma Nakamura, Yasutoshi Ida, Takashi Matsumoto, Takashi Kaburagi

    IASTED Multiconferences - Proceedings of the IASTED International Conference on Artificial Intelligence and Applications, AIA 2013   accepted   51 - 58  2013

     View Summary

    Recent research into time-varying network inference for gene expression data mainly assumes that gene regulatory networks have linear interactions. This assumption is straightforward and requires comparatively simple model building. However, in various previous biological studies, gene expression data have been believed to have nonlinear properties in their regulatory interactions. To address this, we adopted a nonparametric Bayesian regression method (e.g. a Gaussian Process) for predicting interactions into a time-varying network to achieve more flexible regression capability. The proposed method was evaluated on Drosophila melanogaster gene data, which has been used as a benchmark in a number of studies. This dataset, which was measured by a microarray test, is known to include noise. To obtain stronger robustness to noisy data, in our algorithm, we employed the T-Process instead of the conventional Gaussian Process. To the best of our knowledge, this is the first algorithm to apply nonparametric Bayesian regression method to a time-varying gene regulatory network problem. Our basic algorithm employed reversible jump Markov Chain Monte Carlo (RJMCMC) for inference of whole network structures. The method can handle the two inference problems: (i) change point detection and (ii) network structure inference simultaneously.

    DOI

  • Bayesian STAI Anxiety Index Predictions Based on Prefrontal Cortex NIRS Data for the Resting State

    Masakaze Sato, Wakana Ishikawa, Tomohiko Suzuki, Takashi Matsumoto, Takeo Tsujii, Kaoru Sakatani

    OXYGEN TRANSPORT TO TISSUE XXXIV   765   251 - 256  2013

     View Summary

    Several distinctive activity patterns have been observed in the brain at rest. The aim of this study was to determine whether the STAI index can be predicted from changes in the oxy- and deoxy-hemoglobin (Hb) concentrations by using two-channel prefrontal cortex (PFC) NIRS data for the resting state. The study population comprised 19 subjects. Each subject performed four trials, each of which consisted of resting with no task for 3 min. Data were acquired using a portable NIRS device equipped with two channels. The prediction algorithm was derived within a Bayesian machine learning framework. The prediction errors for seven subjects were not greater than 5.0. Because the STAI index varied between 20 and 80, these predictions appeared reasonable. The present method allowed prediction of mental status based on the NIRS data at resting condition obtained in the PFC.

    DOI

  • Nonparametric bayes-based heterogeneous "Drosophila Melanogaster" gene regulatory network inference: T-process regression

    Hiroki Miyashita, Takuma Nakamura, Yasutoshi Ida, Takashi Matsumoto, Takashi Kaburagi

    IASTED Multiconferences - Proceedings of the IASTED International Conference on Artificial Intelligence and Applications, AIA 2013   accepted   51 - 58  2013

     View Summary

    Recent research into time-varying network inference for gene expression data mainly assumes that gene regulatory networks have linear interactions. This assumption is straightforward and requires comparatively simple model building. However, in various previous biological studies, gene expression data have been believed to have nonlinear properties in their regulatory interactions. To address this, we adopted a nonparametric Bayesian regression method (e.g. a Gaussian Process) for predicting interactions into a time-varying network to achieve more flexible regression capability. The proposed method was evaluated on Drosophila melanogaster gene data, which has been used as a benchmark in a number of studies. This dataset, which was measured by a microarray test, is known to include noise. To obtain stronger robustness to noisy data, in our algorithm, we employed the T-Process instead of the conventional Gaussian Process. To the best of our knowledge, this is the first algorithm to apply nonparametric Bayesian regression method to a time-varying gene regulatory network problem. Our basic algorithm employed reversible jump Markov Chain Monte Carlo (RJMCMC) for inference of whole network structures. The method can handle the two inference problems: (i) change point detection and (ii) network structure inference simultaneously.

    DOI

  • Bayesian STAI anxiety index predictions based on prefrontal cortex NIRS data for the resting state

    Masakaze Sato, Wakana Ishikawa, Tomohiko Suzuki, Takashi Matsumoto, Takeo Tsujii, Kaoru Sakatani

    Advances in Experimental Medicine and Biology   765   251 - 256  2013

     View Summary

    Several distinctive activity patterns have been observed in the brain at rest. The aim of this study was to determine whether the STAI index can be predicted from changes in the oxy-and deoxy-hemoglobin (Hb) concentrations by using two-channel prefrontal cortex (PFC) NIRS data for the resting state. The study population comprised 19 subjects. Each subject performed four trials, each of which consisted of resting with no task for 3 min. Data were acquired using a portable NIRS device equipped with two channels. The prediction algorithm was derived within a Bayesian machine learning framework. The prediction errors for seven subjects were not greater than 5.0. Because the STAI index varied between 20 and 80, these predictions appeared reasonable. The present method allowed prediction of mental status based on the NIRS data at resting condition obtained in the PFC. © 2013 Springer Science+Business Media New York.

    DOI PubMed

  • Bayesian event detection for sport games with hidden Markov model

    Shigeru Motoi, Toshie Misu, Yohei Nakada, Tomohiro Yazaki, Go Kobayashi, Takashi Matsumoto, Nobuyuki Yagi

    PATTERN ANALYSIS AND APPLICATIONS   15 ( 1 ) 59 - 72  2012.02

     View Summary

    Event detection can be defined as the problem of detecting when a target event has occurred, from a given data sequence. Such an event detection problem can be found in many fields in science and engineering, such as signal processing, pattern recognition, and image processing. In recent years, many data sequences used in these fields, especially in video data analysis, tend to be high dimensional. In this paper, we propose a novel event detection method for high-dimensional data sequences in soccer video analysis. The proposed method assumes a Bayesian hidden Markov model with hyperparameter learning in addition to the parameter leaning. This is in an attempt to reduce undesired influences from ineffective components within the high-dimensional data. Implemention is performed by Markov Chain Monte Carlo. The proposed method was tested against an event detection problem with sequences of 40-dimensional feature values extracted from real professional soccer games. The algorithm appears functional.

    DOI

  • HANDLING INCOMPLETE MATRIX DATA VIA CONTINUOUS-VALUED INFINITE RELATIONAL MODEL

    Tomohiko Suzuki, Takuma Nakamura, Yasutoshi Ida, Takashi Matsumoto

    ICASSP 2012     2153 - 2156  2012

    DOI

  • HDP-MRF: A Hierarchical Nonparametric Model for Image Segmentation

    Takuma Nakamura, Tatsuhiro Harada, Tomohiko Suzuki, Takashi Matsumoto

    ICPR 2012    2012

  • 0.93nmスペクトル解像度ハイパースペクトラルDatacube生体認証の試み

    赤沢史嗣, 村松大吾, 佐藤優太, 松本 隆, 中村 厚, 宗田孝之

    電子情報通信学会 バイオメトリクス研究会     29 - 38  2012

  • 前頭前野の安静時脳血流酸素代謝の左右非対称性による不安心理状態の推定法

    石川わかな, 福田行克, 松本隆, 竹本尚大, 辻井岳雄, 酒谷薫

    第16回酸素ダイナミクス研究会     33 - 35  2012

  • ノンパラメトリックベイジアンT課程アルゴリズムによる時間的構造変化を考慮した遺伝子発現ネットワーク推定

    宮下弘樹, 中村拓磨, 井田安俊, 鈴木知彦, 松本隆, 鏑木崇史

    情報処理学会 BIO32-MPS91合同発表会    2012

  • HANDLING INCOMPLETE MATRIX DATA VIA CONTINUOUS-VALUED INFINITE RELATIONAL MODEL

    Tomohiko Suzuki, Takuma Nakamura, Yasutoshi Ida, Takashi Matsumoto

    ICASSP 2012     2153 - 2156  2012

    DOI

  • HDP-MRF: A Hierarchical Nonparametric Model for Image Segmentation

    Takuma Nakamura, Tatsuhiro Harada, Tomohiko Suzuki, Takashi Matsumoto

    ICPR 2012    2012

  • HANDLING INCOMPLETE MATRIX DATA VIA CONTINUOUS-VALUED INFINITE RELATIONAL MODEL

    Tomohiko Suzuki, Takuma Nakamura, Yasutoshi Ida, Takashi Matsumoto

    ICASSP 2012    2012

    DOI

  • Online Brain-Machine Co-learning for SSVEP Covert-selective Attention

    Y.Dobashi, C.Serizawa, T.Matsumoto

    Workshop on Tools for Brain Computer Interaction    2012

  • A Bayesian Prediction of STAI at Resting State from 2-channel Prefrontal Cortex NIRS Data

    M.Sato, W.Ishikawa, T.Suzuki, T.Matsumoto, T.Tsujii, K.Sakatani

    JSMBE 51    2012

  • New Method of Analysing NIRS Data from Prefrontal Cortex at Rest

    W.Ishikawa, M.Sato, Y.Fukuda, T.Matsumoto, N.Takemoto, T.Tsujii, K.Sakatani

    ISOTT 2012    2012

  • HDP-MRF: A Hierarchical Nonparametric Model for Image Segmentation

    Takuma Nakamura, Tatsuhiro Harada, Tomohiko Suzuki, Takashi Matsumoto

    ICPR 2012   accepted  2012

  • 0.93nmスペクトル解像度ハイパースペクトラルDatacube生体認証の試み

    赤沢史嗣, 村松大吾, 佐藤優太, 松本 隆, 中村 厚, 宗田孝之

    電子情報通信学会 バイオメトリクス研究会    2012

  • 前頭前野の安静時脳血流酸素代謝の左右非対称性による不安心理状態の推定法

    石川わかな, 福田行克, 松本隆, 竹本尚大, 辻井岳雄, 酒谷薫

    第16回酸素ダイナミクス研究会    2012

  • 前頭前野の安静時NIRS信号による不安心理状態の推定法

    石川わかな, 福田行克, 松本隆, 竹本尚大, 辻井岳雄, 酒谷薫

    第18回日本脳神経モニタリング学会    2012

  • Automatic determination of stopping time of training phase in SSVEP-based brain-machine interface with Bayesian sequential learning

    Yumi Dobashi, Atsushi Takemoto, Shu Shigezumi, Takumi Shiraki, Katsuki Nakamura, Takashi Matsumoto

    Proceedings of the 9th IASTED International Conference on Biomedical Engineering, BioMed 2012     29 - 36  2012

     View Summary

    This paper proposes an EEG-based Brain-Machine Interface (BMI) system such that 1) the machine can determine when to end the learning phase automatically by monitoring the learning progress using the Sequential Error Rate (SER) as an evaluation index and 2) it involves sequential learning in both the brain and the machine in a cooperative manner. In the proposed 'Brain-Machine Co-learning', subjects learn how to use the system by means of real-time visual feedback, whereas the machine learns the subjects' EEG signals by Bayesian sequential learning. The SER refers to the average classification error rate windowed over a short time period, and it represents the status of Bayesian sequential learning in real time. In our proposed approach, subjects can use the system while eliminating unnecessary training. The proposed system was tested against an SSVEP classification problem. The training phase varied for each subject and was sometimes short, yet satisfactory, leading to high classification accuracy.

    DOI

  • HANDLING INCOMPLETE MATRIX DATA VIA CONTINUOUS-VALUED INFINITE RELATIONAL MODEL

    Tomohiko Suzuki, Takuma Nakamura, Yasutoshi Ida, Takashi Matsumoto

    ICASSP 2012    2012

    DOI

  • Online Brain-Machine Co-learning for SSVEP Covert-selective Attention

    Y.Dobashi, C.Serizawa, T.Matsumoto

    Workshop on Tools for Brain Computer Interaction    2012

  • A Bayesian Prediction of STAI at Resting State from 2-channel Prefrontal Cortex NIRS Data

    M.Sato, W.Ishikawa, T.Suzuki, T.Matsumoto, T.Tsujii, K.Sakatani

    JSMBE 51    2012

  • New Method of Analysing NIRS Data from Prefrontal Cortex at Rest

    W.Ishikawa, M.Sato, Y.Fukuda, T.Matsumoto, N.Takemoto, T.Tsujii, K.Sakatani

    ISOTT 2012    2012

  • HDP-MRF: A Hierarchical Nonparametric Model for Image Segmentation

    Takuma Nakamura, Tatsuhiro Harada, Tomohiko Suzuki, Takashi Matsumoto

    ICPR 2012   accepted  2012

  • Image-based Authentication Methods

    T. Matsumoto, A.Nakamura, D.Muramatsu, T.Sota

    2nd World Congress on Forensics 2011     37  2011

  • Image-based Authentication Methods

    T. Matsumoto, A.Nakamura, D.Muramatsu, T.Sota

    2nd World Congress on Forensics 2011     37  2011

  • ベイズ隠れマルコフモデルを用いたスポーツイベント検出の高度化

    矢崎智浩, 本井滋, 小林剛, 松本隆, 三須俊枝, 八木伸行, 中田洋平

    画像ラボ     9 - 16  2011

  • 脳波:SSVEP4クラス判別問題の遂次型学習:Sequential Monte Carlo実装

    重住周, 原英之, 竹本篤史, 土橋由実, 中村克樹, 松本隆

    電子情報通信学会 信学技報   110 ( 460 ) 125 - 130  2011

  • ダイナミック・ガウス過程遺伝子発現ネットワーク 予測:MCMC実装

    菊地貴彰, 鈴木知彦, 中田洋平, 鏑木崇史, 松本隆, 君和田友美, 和田圭司

    情報処理学会研究報告   2011-BIO-25 ( 6 ) 1 - 8  2011

  • Bayesian SSVEP/NIRS-based Brain Signal Decoding with Monte Carlo Implementations

    T. Matsumoto

    MLSP 2011 (key note)    2011

  • Image-based Authentication Methods

    T. Matsumoto, A.Nakamura, D.Muramatsu, T.Sota

    WCF 2011     37  2011

  • Bayesian STAI Anxiety Index Predictions Based On Prefrontal Cortex NIRS Data for the Resting State

    Masakaze Sato, Wakana Ishikawa, Tomohiko Suzuki, Takashi Matsumoto, Takeo Tsujii, Kaoru Sakatani

    ISOTT 2011    2011

  • Infinite Mixture Model Approach for Protein Function Prediction Algorithm Utilizing Hidden Markov Model and Bayesian Network Model with Dirichlet Process Prior

    Takashi Kaburagi, Yukihiro Koizumi, Go Kobayashi, Kousuke Oota, Yohei Nakada, Takashi Matsumoto

    ISMB/ECCB 2011    2011

  • 半教師あり学習を用いたノンパラメトリックトピックモデルによるweb文書の評判分析

    井田安俊, 鈴木知彦, 松本

    IBIS2011    2011

  • 文書情報を考慮した無限関係モデルによるネットワーク推定

    鈴木知彦, 井田安俊, 松本隆

    IBIS2011    2011

  • 可視光/NIR帯域1nm\lambda解像度大量画像データからの個人特徴抽出

    村松大吾, 赤沢史嗣, 白土聡, 松本隆, 中村厚, 宗田孝之

    第1回バイオメトリクスと認識・認証シンポジウム    2011

  • Bayesian SSVEP/NIRS-based Brain Signal Decoding with Monte Carlo Implementations

    T. Matsumoto

    MLSP 2011 (key note)    2011

  • Image-based Authentication Methods

    T. Matsumoto, A.Nakamura, D.Muramatsu, T.Sota

    WCF 2011     37  2011

  • Bayesian STAI Anxiety Index Predictions Based On Prefrontal Cortex NIRS Data for the Resting State

    Masakaze Sato, Wakana Ishikawa, Tomohiko Suzuki, Takashi Matsumoto, Takeo Tsujii, Kaoru Sakatani

    ISOTT 2011    2011

  • Infinite Mixture Model Approach for Protein Function Prediction Algorithm Utilizing Hidden Markov Model and Bayesian Network Model with Dirichlet Process Prior

    Takashi Kaburagi, Yukihiro Koizumi, Go Kobayashi, Kousuke Oota, Yohei Nakada, Takashi Matsumoto

    ISMB/ECCB 2011    2011

  • Visual-based online signature verification using features extracted from video

    Kumiko Yasuda, Daigo Muramatsu, Satoshi Shirato, Takashi Matsumoto

    JOURNAL OF NETWORK AND COMPUTER APPLICATIONS   33 ( 3 ) 333 - 341  2010.05

     View Summary

    We propose a visual-based online signature verification system. The input module of the system consists of only low-cost cameras (webcams) and does not need an electronic tablet. Online signature data are obtained from the images captured by the webcams by tracking the pen tip. The pen tip tracking is implemented by the sequential Monte Carlo method. Then, the distance between the input signature data and reference signature data enrolled in advance is computed. Finally, the input signature is classified as genuine or a forgery by comparing the distance with a threshold. In this paper, we consider seven camera positions. We performed experiments using a private database consisting of 150 genuine signatures to decide the best camera position. The experimental results show that we should place the webcam to the side of the hand. Moreover, we evaluated the proposed system with a camera placed to the side of the hand against a different database consisting of 390 genuine signatures and 1560 skilled forged signatures. The proposed system achieved an equal error rate of 4.1% against this database. (C) 2009 Elsevier Ltd. All rights reserved.

    DOI CiNii

  • Person-Independent Face Localization Under Expression Changes with DP-EM Clustering

    M.Hara, T.Tokuda, A.Matsui, S.Clippingdale, T.Matsumoto

    SPPRA2010     290 - 296  2010

  • Bayes学習のMonte Carlo実装

    松本隆

    映情学技報   34 ( 11 )  2010

  • 混合正規分布推定問題に対するStick-Breaking過程EMアルゴリズム

    鈴木奈緒美, 徳田貴昭, 中田洋平, 松本隆

    電子情報通信学会総合大会   ISS-P-112   12  2010

  • 遮蔽を考慮したCross Entropyに基づく尤度計算によるリアルタイム複数物体追跡

    上條秀一, 宮島雄一, 松井淳, 中田洋平, 松本隆

    電子情報通信学会総合大会   ISS-P-103   3  2010

  • ソフトウェア無線機用離散時間フィルタに関する研究

    峯藤健司, 菅真吾, 松本隆

    電子情報通信学会総合大会   ISS-P-110   10  2010

  • 脳波motor imageryデータのBayes的有効特徴量自動抽出

    土橋由実, 八木佑太圭, 原英之, 中田洋平, 松本隆, 竹本篤史, 中村克樹

    電子情報通信学会総合大会   ISS-P-149   49  2010

  • 脳波SSVEP非線形判別:Bayes学習によるMonte Carlo実装

    原英之, 松本隆, 中村克樹, 竹本篤史, 中田洋平

    電子情報通信学会総合大会   ISS-P-304   135  2010

  • ディリクレ過程EMアルゴリズムを用いた混合離散連続同時確率分布推定

    谷戸崇紀, 徳田貴明, 能鹿島武志, 松本隆

    電子情報通信学会総合大会   ISS-P-111   11  2010

  • ColorSIFT特徴点を用いたディリクレ過程混合モデルによる画像分類

    下田康夫, 能鹿島武志, 徳田貴昭, 松本隆

    電子情報通信学会総合大会   ISS-P-113   13  2010

  • Monte Carlo based HMM によるMotor Imagery判別

    八木佑太圭, 中田洋平, 松本隆, 竹本篤史, 中村克樹

    電子情報通信学会総合大会   ISS-P-102   2  2010

  • 文字入力支援ソフトDasher日本語実装の実験

    鈴木知彦, 鏑木崇史, Mackay David, 松本隆

    電子情報通信学会総合大会   ISS-P-115   15  2010

  • 環境情報へのベイジアンネットワーク的なアプローチ

    永澤惇, 徳田貴昭, 中田洋平, 鏑木崇史, 松本隆

    電子情報通信学会総合大会   ISS-P-146   46  2010

  • 顔特徴点抽出アルゴリズムのリアルタイム実装の試み

    斉藤千晶, 原美咲, 松井淳, 松本隆

    電子情報通信学会総合大会   ISS-P-256   124  2010

  • クロスエントロピーを用いた複数移動物体追跡の高速化

    上條秀一, 宮島雄一, 松井淳, 中田洋平, 松本隆

    情報処理学会研究報告   2010-CVIM-171 ( 9 )  2010

  • ベイズ隠れマルコフモデルを用いたスポーツイベント検出の高精度化

    矢崎智浩, 三須俊枝, 中田洋平, 本井滋, 小林剛, 松本隆, 八木伸行

    電子情報通信学会信学技報 HIP   109 ( 471 ) 401 - 406  2010

  • ディリクレ過程事前分布を用いた一般物体認識のための確率生成モデルの拡張と推定法

    能鹿島武志, 徳田貴昭, 中田洋平, 松本隆

    画像の認識・理解シンポジウム(MIRU)2010    2010

  • A Gene Regulatory Network prediction algorithm Using Gaussian Network model with Box-Cox transformation

    H.Miyachika, J.Maruyama, T.Kaburagi, Y.Nakada, T.Matsumoto, T.Kimiwada, K.Wada

    ISMB2010    2010

  • A non-parametric Bayesian algorithm for predicting gene regulatory networks with a Gaussian process

    T.Kikuchi, Y.Nakada, T.Kaburagi, T.Matsumoto

    ISMB2010    2010

  • A Novel Gene Ontology Prediction Algorithm Using Infinite Mixtures of Hidden Markov and Binary Models with A Dirichlet Process Prior

    Takashi Kaburagi, Natsumi Tagoto, Kousuke Oota, Takaaki Tokuda, Yohei Nakada, Takashi Matsumoto

    ISMB2010    2010

  • CAMERA-BASED ONLINE SIGNATURE VERIFICATION SYSTEM: EFFECTS OF CAMERA POSITIONS

    Satoshi Shirato, Daigo Muramatsu, Takashi Matsumoto

    WAC 2010    2010

  • 脳波SSVEP2値判別問題における逐次誤差率の評価~ベイズ的逐次型学習によるSequential Monte Carlo実装~

    原英之, 竹本篤史, 土橋由実, 中村克樹, 松本隆

    電子情報通信学会信学技報 MBE2010-31   110 ( 226 ) 17 - 22  2010

  • 背景を考慮したCross Entropyによる高速な複数物体追跡

    上條秀一, 宮島雄一, 松井淳, 中田洋平, 村松大吾, 松本隆

    画像電子学会誌   39 ( 5 ) 571 - 579  2010

    DOI CiNii

  • 2次 sinc 関数特性を持つチャージサンプリングフィルタの低消費電力化・小面積化の検討

    峯藤健司, 松本隆

    電子情報通信学会 信学技報   110 ( 344 ) 65 - 70  2010

  • Person-Independent Face Localization Under Expression Changes with DP-EM Clustering

    M.Hara, T.Tokuda, A.Matsui, S.Clippingdale, T.Matsumoto

    SPPRA2010     290 - 296  2010

  • Monte Carlo-based mouse nuclear receptor superfamily gene regulatory network prediction: Stochastic dynamical system on graph with Zipf prior

    Yusuke Kitamura, Tomomi Kimiwada, Jun Maruyama, Takashi Kaburagi, Takashi Matsumoto, Keiji Wada

    IPSJ Transactions on Bioinformatics   3   24 - 39  2010

     View Summary

    A Monte Carlo based algorithm is proposed to predict gene regulatory network structure of mouse nuclear receptor superfamily, about which little is known although those genes are believed to be related with several difficult diseases. The gene expression data is regarded as sample vector trajectories from a stochastic dynamical system on a graph. The problem is formulated within a Bayesian framework where the graph prior distribution is assumed to follow a Zipf distribution. Appropriateness of a graph is evaluated by the graph posterior mean. The algorithm is implemented with the Exchange Monte Carlo method. After validation against synthesized data, an attempt is made to use the algorithm for predicting network structure of the target, the mouse nuclear receptor superfamily. Several remarks are made on the feasibility of the predicted network from a biological viewpoint. © 2010 Information Processing Society of Japan.

    DOI CiNii

  • A Gene Regulatory Network prediction algorithm Using Gaussian Network model with Box-Cox transformation

    H.Miyachika, J.Maruyama, T.Kaburagi, Y.Nakada, T.Matsumoto, T.Kimiwada, K.Wada

    ISMB2010    2010

  • A non-parametric Bayesian algorithm for predicting gene regulatory networks with a Gaussian process

    T.Kikuchi, Y.Nakada, T.Kaburagi, T.Matsumoto

    ISMB2010    2010

  • A Novel Gene Ontology Prediction Algorithm Using Infinite Mixtures of Hidden Markov and Binary Models with A Dirichlet Process Prior

    Takashi Kaburagi, Natsumi Tagoto, Kousuke Oota, Takaaki Tokuda, Yohei Nakada, Takashi Matsumoto

    ISMB2010    2010

  • CAMERA-BASED ONLINE SIGNATURE VERIFICATION SYSTEM: EFFECTS OF CAMERA POSITIONS

    Satoshi Shirato, Daigo Muramatsu, Takashi Matsumoto

    WAC 2010    2010

  • Sequential error rate evaluation of SSVEP classification problem with Bayesian sequential learning

    Hideyuki Hara, Atsushi Takemoto, Yumi Dobashi, Katsuki Nakamura, Takashi Matsumoto

    Proceedings of the IEEE/EMBS Region 8 International Conference on Information Technology Applications in Biomedicine, ITAB    2010

     View Summary

    An attempt was made to evaluate the Sequential Error Rate (SER) of an SSVEP classification problem with a Bayesian sequential learning algorithm. Sequential Error Rate refers to the average classification error rate windowed over a short trial period. The algorithm was implemented by the Sequential Monte Carlo method. As opposed to the batch learning algorithm, the sequential learning algorithm does not divide the data into training and test datasets
    rather, it starts learning with the first single trial data and proceeds with the learning sequentially using the rest of the data. The algorithm was tested against an SSVEP classification problem. The algorithm appeared functional © 2010 IEEE.

    DOI

  • ベイズ学習の実装:MCMC/SMC/DPEM

    松本隆

    電子情報通信学会 信学技報   108 ( 484 ) 39 - 42  2009

  • ディリクレ過程事前分布言語モデルに対する事後確率最大化推定法

    徳田貴昭, 木村智明, 中田洋平, 松本隆

    電子情報通信学会 信学技報   108 ( 432 ) 109 - 114  2009

  • ディリクレ過程事前分布EMアルゴリズムによる顔画像検出

    後藤祐, 木村智明, 松井淳, 中田洋平, 松本隆

    電子情報通信学会 信学技報   108 ( 432 ) 37 - 42  2009

  • Semi-supervised learning scheme using Dirichlet process EM-algorithm

    Tomoaki Kimura, Yohei Nakada, Arnaud Doucet, Takashi Matsumoto

    電子情報通信学会 信学技報   108 ( 484 ) 77 - 82  2009

  • Online Signature Verification Based on User-Generic Fusion Model with Markov Chain Monte Carlo, Taking into Account User Individuality

    Kyosuke Koishi, Shintaro Kinoshita, Daigo Muramatsu, Takashi Matsumoto

    J.Advanced Computational Intelligence and Intelligent Informatics   13 ( 4 ) 447 - 456  2009

  • 標的タンパク質の動的構造を考慮したフォーマコフォアモデルの構築と評価

    平山和徳, 小林誠一, 松本隆, 広川貴次

    日本薬学会第129年会、物理系薬学 計算機化学1    2009

  • ソフトウェア無線機向け離散時間フィルタの高次化に関する検討

    西島武良, 松本 隆

    電子情報通信学会総合大会    2009

  • 文字入力支援システムDasherの日本語実装

    鏑木崇史, 福田真啓, 瀬賀一恵, 松本隆, David J. C. MacKay

    情報処理学会全国大会    2009

  • 動画像から抽出した特徴を用いたオンライン署名認証

    安田久美子, 白土聡, 村松大吾, 松本隆

    第16回バイオメトリックシステムセキュリティ研究発表会     9 - 16  2009

  • ウェブカメラを用いたオンライン署名認証システムの実装

    南康文, 白土聡, 安田久美子, 村松大吾, 松本隆

    第16回バイオメトリックシステムソサエティ研究発表会     17 - 24  2009

  • ユーザ個人性を考慮し、AdaBoostを用いたユーザ共通型オンライン署名認証

    小石恭輔, 村松大吾, 松本隆

    第16回バイオメトリックシステムセキュリティ研究発表会     1 - 8  2009

  • Monte Carlo-Based Bayesian Prediction of Gene Regulatory Networks with Zipf Distribution: Mouse Nuclear Receptor Superfamily

    H.Miyachika, Y.Kitamura, T.Kimiwada, J.Maruyama, T.Kaburagi, T.Matsumoto, K.Wada

    ISMB2009    2009

  • A Bayesian Monte Carlo Hidden Markov Model Approach to Transmembrane Protein Structure Prediction

    T. Kaburagi, T. Matsumoto

    ISMB2009    2009

  • Online Signature Verification Algorithm with a User-Specific Global-Parameter Fusion Model

    D.Muramatsu, T.Matsumoto

    IEEE International Conference on Systems, Man and Cybernetics    2009

    DOI

  • Camera-Based Online Signature Verification with Sequential Marginal Likelihood Change Dete

    Daigo Muramatsu, Kumiko Yasuda, Satoshi Shirato, Takashi Matsumoto

    CAIP2009     229 - 236  2009

    DOI

  • Maximum A Posterior Estimation For Dirichlet Process Language Models

    T.Tokuda, T.Kimura, Y.Nakada, T.Matsumoto

    7th Workshop on Bayesian Nonparametrics    2009

  • Dirichlet Process EM algorithm for Semi-supervised Learning

    T.Kimura, Y.Nakada, T.Tokuda, T.Matsumoto

    7th Workshop on Bayesian Nonparametrics    2009

  • ベイズ学習の実装

    松本隆

    電子情報通信学会誌   92 ( 10 ) 853 - 860  2009

  • Signature Recognition

    O.Henniger, D.Muramatsu, T.Matsumoto, I.Yoshimura, M.Yoshimura

    Encyclopedia of Biometrics、Li, Stan Z.(ed),     1196 - 1205  2009

  • Semi-supervised learning scheme using Dirichlet process EM-algorithm

    Tomoaki Kimura, Yohei Nakada, Arnaud Doucet, Takashi Matsumoto

    電子情報通信学会 信学技報   108 ( 484 ) 77 - 82  2009

  • Online Signature Verification Based on User-Generic Fusion Model with Markov Chain Monte Carlo, Taking into Account User Individuality

    Kyosuke Koishi, Shintaro Kinoshita, Daigo Muramatsu, Takashi Matsumoto

    J.Advanced Computational Intelligence and Intelligent Informatics   13 ( 4 ) 447 - 456  2009

  • Biometric person authentication method using camera-based online signature acquisition

    Daigo Muramatsu, Kumiko Yasuda, Takashi Matsumoto

    Proceedings of the International Conference on Document Analysis and Recognition, ICDAR     46 - 50  2009

     View Summary

    A camera-based online signature verification system is proposed in this paper. One web camera is used for data acquisition, and a sequential Monte Carlo method is used for tracking a pen tip. Several distances are computed from an online signature, and a fusion model trained by using AdaBoost combines the distances and computes a final score. Preliminary experiments were performed by using a private database. The proposed system yielded an equal error rate (EER) of 4.0%. © 2009 IEEE.

    DOI

  • Monte Carlo-Based Bayesian Prediction of Gene Regulatory Networks with Zipf Distribution: Mouse Nuclear Receptor Superfamily

    H.Miyachika, Y.Kitamura, T.Kimiwada, J.Maruyama, T.Kaburagi, T.Matsumoto, K.Wada

    ISMB2009    2009

  • A Bayesian Monte Carlo Hidden Markov Model Approach to Transmembrane Protein Structure Prediction

    T. Kaburagi, T. Matsumoto

    ISMB2009    2009

  • Online Signature Verification Algorithm with a User-Specific Global-Parameter Fusion Model

    D.Muramatsu, T.Matsumoto

    IEEE International Conference on Systems, Man and Cybernetics    2009

    DOI

  • Camera-Based Online Signature Verification with Sequential Marginal Likelihood Change Dete

    Daigo Muramatsu, Kumiko Yasuda, Satoshi Shirato, Takashi Matsumoto

    CAIP2009     229 - 236  2009

    DOI

  • Maximum A Posterior Estimation For Dirichlet Process Language Models

    T.Tokuda, T.Kimura, Y.Nakada, T.Matsumoto

    7th Workshop on Bayesian Nonparametrics    2009

  • Dirichlet Process EM algorithm for Semi-supervised Learning

    T.Kimura, Y.Nakada, T.Tokuda, T.Matsumoto

    7th Workshop on Bayesian Nonparametrics    2009

  • Signature Recognition

    O.Henniger, D.Muramatsu, T.Matsumoto, I.Yoshimura, M.Yoshimura

    Encyclopedia of Biometrics、Li, Stan Z.(ed),     1196 - 1205  2009

  • 自然逐次事前分布によるオンラインベイズ学習

    中田洋平, 若原牧生, 松本隆

    電子情報通信学会論文誌 A,   J91-A ( 2 ) 243 - 259  2008

  • A Novel Hierarchical Bayesian HMM For Multi-Dimensional Discrete Data

    S. Motoi, Y. Nakada, T. Misu, T. Matsumoto, N. Yagi

    28th IASTED Conference on Artificial Intelligence and Applications    2008

  • Monte Carlo SLAM method for a Small Mobile Robot with Short-Range Sensors

    K. Yamada, Y. Nakada, T. Matsumoto

    27th IASTED International Conference on Modelling, Identification, and Control    2008

  • A Generalized Hidden Markov Model Approach to Transmembrane Region Prediction with Poisson Distribution as State Duration Probabilities

    Kaburagi Takashi, Matsumoto Takashi

    ipsjdc   4   193 - 206  2008

     View Summary

    We present a novel algorithm to predict transmembrane regions from a primary amino acid sequence. Previous studies have shown that the Hidden Markov Model (HMM) is one of the powerful tools known to predict transmembrane regions; however, one of the conceptual drawbacks of the standard HMM is the fact that the state duration, i.e., the duration for which the hidden dynamics remains in a particular state follows the geometric distribution. Real data, however, does not always indicate such a geometric distribution. The proposed algorithm utilizes a Generalized Hidden Markov Model (GHMM), an extension of the HMM, to cope with this problem. In the GHMM, the state duration probability can be any discrete distribution, including a geometric distribution. The proposed algorithm employs a state duration probability based on a Poisson distribution. We consider the two-dimensional vector trajectory consisting of hydropathy index and charge associated with amino acids, instead of the 20 letter symbol sequences. Also a Monte Carlo method (Forward/Backward Sampling method) is adopted for the transmembrane region prediction step. Prediction accuracies using publicly available data sets show that the proposed algorithm yields reasonably good results when compared against some existing algorithms.

    DOI CiNii

  • バンドパスチャージサンプリングの単相入力化

    T. Murakami, T. Matsumoto

    電子情報通信学会 第15回シリコンアナログRF研究会   RF2007 ( 4 )  2008

  • GibbsBoost 顔検出と映像監視業務への応用

    松井淳, 後藤祐, 木村彰夫, 中田洋平, 松本隆, クリッピングデルサイモン, 藤井真人, 八木伸行

    映像情報メディア学会誌   62 ( 3 ) 408 - 413  2008

    DOI CiNii

  • ベイズ的手法による動画像顔検出の高速化と高精度化

    松井淳, サイモン クリピングデル, 八木伸行, 松本隆

    電子情報通信学会2008年総合大会論文集   D-12-26   157  2008

  • 未知非線形システムに対するオンライン変化検出手法:逐次モンテカルロ法を用いたベイズ的アプローチ

    中田洋平, 松本隆

    電子情報通信学会論文誌 A   J91-A ( 6 ) 654 - 668  2008

  • Backward Smoothing Approach to Transmembrane Protein Structure Prediction with Stochastic Dynamical Systems

    Takashi Kaburagi, Takashi Matsumoto

    J. Computational Intelligence in Bioinformatics   1 ( 1 ) 13 - 33  2008

  • ヒルクライミング法を用いた攻撃に強いオンライン署名認証アルゴリズム

    村松大吾, 厳 維娜, 松本 隆

    電子情報通信学会論文誌A   J91-A ( 10 ) 983 - 988  2008

  • Bayesian Sequential Face Detection with Automatic Re-initialization

    Atsushi Matsui, Simon Clippingdale, Takashi Matsumoto

    International Conference on Pattern Recognition    2008

  • Identification of novel inhibitors for ubiquitin C-terminal hydrolase-L3 by virtual screening

    K. Hirayama, S. Aoki, K. Nishikawa, T. Matsumoto, K. Wada

    16th Annual Int'l Conference Intelligent Systems for Molecular Biology    2008

    DOI

  • An Approach to Predict Transmembrane Protein Structure with Stochastic Dynamical Systems Using Backward Smoothing Scheme

    T. Kaburagi, T. Matsumoto

    16th Annual Int'l Conference Intelligent Systems for Molecular Biology    2008

  • 神経疾患におけるLAMP-2サブタイプ特異的な発現調節機構

    藤本 陽平, 古田 晶子, 松本 隆, 和田 圭司

    第31回日本神経科学大会    2008

  • GibbsBoost顔検出と映像監視業務への応用

    松井淳, 後藤祐, 木村彰夫, 中田洋平, 松本隆, クリッピングデルサイモン, 藤井真人, 八木伸行

    画像ラボ   19 ( 8 ) 16 - 21  2008

    CiNii

  • A Generalized Hidden Markov Model Approach to Transmembrane Region Prediction with Poisson Distribution as State Duration Probabilities

    T. Kaburagi, T. Matsumoto

    情報処理学会第14回バイオ情報学研究発表会    2008

  • A Hierarchical Bayesian Hidden Markov Model for Multi-Dimensional Discrete Data

    Shigeru Motoi, Yohei Nakada, Toshie Misu, Tomohiro Yazaki, Takashi Matsumoto, Nobuyuki Yagi

    In-Tech Publications     354 - 374  2008

  • Visual-based Online Signature Verification by Pen Tip Tracking

    Kumiko Yasuda, Daigo Muramatsu, Takashi Matsumoto

    Int’l Conference on Computational Intelligence for Modelling, Control and Automation    2008

    DOI

  • Online Signature Verification based on User-Generic Model with Markov Chain Monte Carlo, Taking into Account User Individuality

    Kyosuke Koishi, Shintaro Kinoshita, Daigo Muramatsu, Takashi Matsumoto

    SCIS & ISIS 2008     1635 - 1640  2008

  • Virtual screening によるUCH-L3 の新規阻害剤同定

    平山和徳, 青木俊介, 西川香里, 松本隆, 和田圭司

    ゲノム科学若手研究会(Bioinformatics研究 若手の会)    2008

  • Identification of novel inhibitors for ubiquitin C-terminal hydrolase-L3 by virtual screening

    K. Hirayama, S. Aoki, K. Nishikawa, T. Matsumoto, K. Wada

    The 2nd Taiwan-Japan Young Researchers Conference on Computational and Systems Biology     45  2008

  • Hidden Occlusion Variable Approach to Multiple Object Tracking with Sequential Monte Carlo Implementation

    Shuichi Kamijo, Yuichi Miyajima, Atsushi Matsui, Yohei Nakada, Takashi Matsumoto

    KJPR2008     69 - 70  2008

  • Face Localization under Expression Changes Based on Deformable Template Matching with Clustering

    Misaki Hara, Atsushi Matsui, Takashi Matsumoto

    KJPR2008     93 - 94  2008

  • Bayesian Video Face Detection with Applications in Broadcasting

    Atsushi Matsui, Simon Clippingdale, Norifumi Okabe, Takashi Matsumoto

    KJPR2008     95 - 96  2008

  • A Novel Hierarchical Bayesian HMM For Multi-Dimensional Discrete Data

    S. Motoi, Y. Nakada, T. Misu, T. Matsumoto, N. Yagi

    28th IASTED Conference on Artificial Intelligence and Applications    2008

  • Monte Carlo SLAM method for a Small Mobile Robot with Short-Range Sensors

    K. Yamada, Y. Nakada, T. Matsumoto

    27th IASTED International Conference on Modelling, Identification, and Control    2008

  • A Generalized Hidden Markov Model Approach to Transmembrane Region Prediction with Poisson Distribution as State Duration Probabilities

    T.Kaburagi, T.Matsumoto

    IPSJ Journal   4   193 - 206  2008

  • Backward Smoothing Approach to Transmembrane Protein Structure Prediction with Stochastic Dynamical Systems

    Takashi Kaburagi, Takashi Matsumoto

    J. Computational Intelligence in Bioinformatics   1 ( 1 ) 13 - 33  2008

  • Bayesian Sequential Face Detection with Automatic Re-initialization

    Atsushi Matsui, Simon Clippingdale, Takashi Matsumoto

    International Conference on Pattern Recognition    2008

  • Identification of novel inhibitors for ubiquitin C-terminal hydrolase-L3 by virtual screening

    K. Hirayama, S. Aoki, K. Nishikawa, T. Matsumoto, K. Wada

    16th Annual Int'l Conference Intelligent Systems for Molecular Biology    2008

    DOI

  • An Approach to Predict Transmembrane Protein Structure with Stochastic Dynamical Systems Using Backward Smoothing Scheme

    T. Kaburagi, T. Matsumoto

    16th Annual Int'l Conference Intelligent Systems for Molecular Biology    2008

  • A Generalized Hidden Markov Model Approach to Transmembrane Region Prediction with Poisson Distribution as State Duration Probabilities

    T. Kaburagi, T. Matsumoto

    情報処理学会第14回バイオ情報学研究発表会    2008

  • A Hierarchical Bayesian Hidden Markov Model for Multi-Dimensional Discrete Data

    Shigeru Motoi, Yohei Nakada, Toshie Misu, Tomohiro Yazaki, Takashi Matsumoto, Nobuyuki Yagi

    In-Tech Publications     354 - 374  2008

  • ONLINE BAYESIAN LEARNING FOR DYNAMICAL CLASSIFICATION PROBLEM USING NATURAL SEQUENTIAL PRIOR

    Kazue Sega, Yohei Nakada, Takashi Matsumoto

    2008 IEEE WORKSHOP ON MACHINE LEARNING FOR SIGNAL PROCESSING     392 - 397  2008

     View Summary

    Classification problems in dynamical environments are in many fields, including signal processing and pattern recognition. In this paper, we propose a novel Bayesian approach to classification in a dynamical environment. The proposed approach employs natural sequential prior to improve online learning for an online classifier model. By using the natural sequential prior, the proposed approach describes the dynamical changes in the classifier model's parameters in a more natural manner. For comparison, the proposed approach and a conventional approach are validated by means of several numerical experiments.

    DOI

  • Visual-based Online Signature Verification by Pen Tip Tracking

    Kumiko Yasuda, Daigo Muramatsu, Takashi Matsumoto

    Int’l Conference on Computational Intelligence for Modelling, Control and Automation    2008

    DOI

  • Online Signature Verification based on User-Generic Model with Markov Chain Monte Carlo, Taking into Account User Individuality

    Kyosuke Koishi, Shintaro Kinoshita, Daigo Muramatsu, Takashi Matsumoto

    SCIS & ISIS 2008     1635 - 1640  2008

  • Identification of novel inhibitors for ubiquitin C-terminal hydrolase-L3 by virtual screening

    K. Hirayama, S. Aoki, K. Nishikawa, T. Matsumoto, K. Wada

    The 2nd Taiwan-Japan Young Researchers Conference on Computational and Systems Biology     45  2008

  • Hidden Occlusion Variable Approach to Multiple Object Tracking with Sequential Monte Carlo Implementation

    Shuichi Kamijo, Yuichi Miyajima, Atsushi Matsui, Yohei Nakada, Takashi Matsumoto

    KJPR2008     69 - 70  2008

  • Face Localization under Expression Changes Based on Deformable Template Matching with Clustering

    Misaki Hara, Atsushi Matsui, Takashi Matsumoto

    KJPR2008     93 - 94  2008

  • Bayesian Video Face Detection with Applications in Broadcasting

    Atsushi Matsui, Simon Clippingdale, Norifumi Okabe, Takashi Matsumoto

    KJPR2008     95 - 96  2008

  • Identification of novel chemical inhibitors for ubiquitin C-terminal hydrolase-L3 by virtual screening

    Kazunori Hirayama, Shunsuke Aoki, Kaori Nishikawa, Takashi Matsumoto, Keiji Wada

    BIOORGANIC & MEDICINAL CHEMISTRY   15 ( 21 ) 6810 - 6818  2007.11

     View Summary

    UCH-L3 (ubiquitin C-terminal hydrolase-L3) is a de-ubiquitinating enzyme that is a component of the ubiquitin-proteasome system and known to be involved in programmed cell death. A previous study of high-throughput drug screening identified an isatin derivative as a UCH-L3 inhibitor. In this study, we attempted to identify a novel inhibitor with a different structural basis. We performed in silico structure-based drug design (SBDD) using human UCH-L3 crystal structure data (PDB code; IXD3) and the virtual compound library (ChemBridge CNS-Set), which includes 32,799 chemicals. By a two-step virtual screening method using DOCK software (first screening) and GOLD software (second screening), we identified 10 compounds with GOLD scores of over 60. To address whether these compounds exhibit an inhibitory effect oil the de-ubiquitinating activity of UCH-L3, we performed all enzymatic assay using ubiquitin-7-amido-4-methylcoumarin (Ub-AMC) as the substrate. As a result, we identified three compounds with similar basic dihydro-pyrrole skeletons as UCH-L3 inhibitors. These novel compounds may be useful for the research of UCH-L3 function, and in drug development for UCH-L3-associated diseases. (C) 2007 Elsevier Ltd. All rights reserved.

    DOI

  • Pruned resampling: Probabilistic model selection schemes for sequential face recognition

    Atsushi Matsui, Simon Clippingdale, Takashi Matsumoto

    IEICE TRANSACTIONS ON INFORMATION AND SYSTEMS   E90D ( 8 ) 1151 - 1159  2007.08

     View Summary

    This paper proposes probabilistic pruning techniques for a Bayesian video face recognition system. The system selects the most probable face model using model posterior distributions, which can be calculated using a Sequential Monte Carlo (SMC) method. A combination of two new pruning schemes at the resampling stage significantly boosts computational efficiency by comparison with the original online learning algorithm. Experimental results demonstrate that this approach achieves better performance in terms of both processing time and ID error rate than a contrasting approach with a temporal decay scheme.

    DOI CiNii

  • Adaptation and change detection with a sequential Monte Carlo scheme

    Takashi Matsumoto, Kuniaki Yosui

    IEEE TRANSACTIONS ON SYSTEMS MAN AND CYBERNETICS PART B-CYBERNETICS   37 ( 3 ) 592 - 606  2007.06

     View Summary

    Given the sequential data from an unknown target system with changing parameters, the first part of this paper discusses online algorithms that adapt to smooth as well as abrupt changes. This paper examines four different parameter/hyperparameter dynamics for online learning and compares their performance within an online Bayesian learning framework. Using the dynamics that performed best in the first part, the second part of this paper attempts to perform change detection in unknown systems in terms of the time dependence of the marginal likelihood. Because of the sequential nature of the algorithms, a sequential Monte Carlo scheme (particle filter) is a natural means for implementation.

    DOI PubMed CiNii

  • Transmembrane structure predictions with hydropathy index/charge two-dimensional trajectories of stochastic dynamical systems

    Takashi Kaburagi, Daigo Muramutsu, Takashi Matsumoto

    Journal of Bioinformatics and Computational Biology   5 ( 3 ) 669 - 692  2007.06

     View Summary

    A novel algorithm is proposed for predicting transmembrane protein secondary structure from two-dimensional vector trajectories consisting of a hydropathy index and formal charge of a test amino acid sequence using stochastic dynamical system models. Two prediction problems are discussed. One is the prediction of transmembrane region counts
    another is that of transmembrane regions, i.e. predicting whether or not each amino acid belongs to a transmembrane region. The prediction accuracies, using a collection of well-characterized transmembrane protein sequences and benchmarking sequences, suggest that the proposed algorithm performs reasonably well. An experiment was performed with a glutamate transporter homologue from Pyrococcus horikoshii. The predicted transmembrane regions of the five human glutamate transporter sequences and observations based on the computed likelihood are reported. © Imperial College Press.

    DOI PubMed

  • BAYESIAN ONLINE CLASSIFICATION USING RAO-BLACKWELLISED SMC AND ITS APPLICATION

    T. Kudo, Y. Nakada, T. Matsumoto

    SPPRA 2007     222 - 227  2007

  • ベイズ学習に基づく隠れマルコフモデルを用いたスポーツ映像解析におけるイベント検出

    本井滋, 三須俊彦, 中田洋平, 松本隆, 八木伸行

    情報処理学会 コンピュータビジョンとイメージメディア研究会   157 ( 18 ) 133 - 139  2007

  • Bayesian Angle Information HMM with a von Mises Distribution and its Implementation using a Bayesian Monte Carlo Method

    H. Sasaki, Y. Nakada, T. Kaburagi, T. Matsumoto

    Proc. European Symposium on Time Series Prediction     29 - 38  2007

  • ユーザ共通Fusionモデルを用いたオンライン署名認証

    村松大吾, 本郷保範, 松本隆

    電子情報通信学会論文誌 D   J90-D ( 2 ) 450 - 459  2007

  • ユーザ共通Fusionモデルを用いたオンライン署名認証手法における個人性の検討

    村松大吾, 木下伸太朗, 松本隆

    2007年 暗号と情報セキュリティシンポジウム (SCIS2007)    2007

  • 松井淳、後藤祐、木村彰夫、中田洋平、松本隆、サイモン・クリッピンデル、藤井真人、八木伸行

    GibbsBoost顔検出を用いた放送映像中の人物遮蔽状態の自動検知

    DIA2007 動的画像処理実利用化ワークショップ     278 - 281  2007

  • Online Bayesian modeling and prediction of nonlinear systems - Sequential Monte Carlo approach

    T. Matsumoto

    METHODS OF INFORMATION IN MEDICINE   46 ( 2 ) 96 - 101  2007

     View Summary

    Objectives: Given time-series data from an unknown target system, one often wants to build a model for the system behind the data and make predictions. If the target system can be assumed to be linear, there are means of modeling and predicting the target system in question. If, however, one cannot assume the system is linear, various linear theories have natural limitations in terms of modeling and predictive capabilities. This paper attempts to construct a model from time-series data and make an online prediction when the linear assumption is not valid.
    Methods: The problem is formulated within a Bayesian framework implemented by the Sequential Monte Carlo method. Online Bayesian learning/prediction requires computation of a posterior distribution in a sequential manner as each datum arrives. The Sequential Monte Carlo method computes the importance weight in order to draw samples from the posterior distribution. The scheme is tested against time-series data from a noisy Rossler system.
    Results: The test time-series data is the x-coordinate of the trajectory generated by a noisy Roessler system. Attempts are made with regard to online reconstruction of the attractor and online prediction of the time-series data.
    Conclusions: The proposed algorithm appears to be functional. The algorithm should be tested against real world data.

  • オンラインベイズ学習による混合数未知の多次元混合正規分布推定

    伊藤慶太, 中田洋平, 松本隆

    平成19年電気学会全国大会論文集   3   135 - 136  2007

  • 隠れ変数を考慮したBBSMC法による自己位置地図同時推定

    山田康平, 中田洋平, 松本隆

    2007年電子情報通信学会総合大会論文集   D-12-64   180  2007

  • 有限差分確率近似法を用いたメタ強化学習

    関川昭二, 中田洋平, 松本隆

    2007年電子情報通信学会全国大会論文集   D-8-10   103  2007

  • 署名認証における人工偽筆生成と,それを用いたモデル学習

    加藤雄大, 村松大吾, 松本隆

    第9回ユビキタスネットワーク社会におけるバイオメトリクスセキュリティ研究会     34 - 39  2007

  • オンライン署名認証における異なる種類の偽筆に対する認証精度評価

    木下伸太朗, 村松大吾, 松本隆

    第9回ユビキタスネットワーク社会におけるバイオメトリクスセキュリティ研究会     48 - 51  2007

  • バーチャルスクリーニングによるUCH-L3低分子阻害剤の新規リード化合物の同定

    平山和徳, 青木俊介, 松本隆, 和田圭司

    第80回日本薬理学会年会 中枢神経   9  2007

  • Effectiveness of Pen Pressure, Azimuth, and Altitude Features for Online Signature Verification

    Daigo Muramatsu, Takashi Matsumoto

    2nd International Conference on Biometrics, Lecture Notes in Computer Science   4642   503 - 512  2007

    DOI

  • モンテカルロ法を用いたオンライン署名認証

    村松大吾, 加藤雄大, 松本隆

    ヒューマンインターフェース学会論文誌   9 ( 2 ) 191 - 200  2007

  • GibbsBoostによる類似文章検索の検討

    山田一郎, 中田洋平, 松井淳, 松本隆, 三浦菊佳, 住吉英樹, 八木伸行

    言語処理学会第13回年次大会発表論文集     538 - 541  2007

  • ベイズ的動画像顔検出における顔候補領域の逐次予測

    松井 淳, サイモン クリピングデル, 松本 隆

    第6回情報科学技術フォーラム(FIT2007)一般講演論文集   3 ( H-025 ) 59 - 60  2007

  • 動的判別問題に対するオンラインベイズ学習への自然逐次事前分布の導入

    瀬賀一恵, 中田洋平, 松本隆

    電子情報通信学会基礎・境界ソサエティ大会   A-2-3   25  2007

  • On-line and Batch Inference for Bioinformatics Data

    T. Matsumoto

    Banff International Resarch Station Workshop on Bioinformatics, Genetics and Stochastic Computation: Bridging the Gap    2007

  • ヒストグラムを用いたピクセル生成モデルにもとづく逐次モンテカルロ動画像追跡

    宮島雄一, 松井淳, 中田洋平, 松本隆

    電子情報通信学会信学技報   PRMU2007 ( 105 ) 75 - 80  2007

  • ユーザ共通Fusionモデルを用いたオンライン署名認証

    村松大吾, 松本隆

    画像ラボ   18 ( 10 ) 7 - 11  2007

    CiNii

  • 顔画像によるオンライン署名認証 Sequentiao Monte Carloを用いたペン先追跡

    安田久美子, 村松大吾, 松井淳, 松本隆

    電子情報通信学会 信学技報   PRMU2007 ( 145 ) 53 - 58  2007

  • Gibbsboost: A boosting algorithm using a sequential Monte Carlo approach

    Yohei Nakadam, Yusuke Mouri, Yasunori Hongo, Takashi Matsumoto

    Proceedings of the 2006 16th IEEE Signal Processing Society Workshop on Machine Learning for Signal Processing, MLSP 2006     259 - 264  2007

     View Summary

    This study proposes a novel boosting algorithm, GibbsBoost. A Gibbs distribution of a weaklearner sequence with a specific loss (energy) function is used in this algorithm as the posterior distribution in Bayesian learning. Weaklearner sequence samples are recursively drawn from the distribution via Sequential Monte Carlo. The predictions are derived from a combination of the weaklearner sequence samples. The proposed algorithm is demonstrated by using a numerical example. © 2006 IEEE.

    DOI

  • BAYESIAN ONLINE CLASSIFICATION USING RAO-BLACKWELLISED SMC AND ITS APPLICATION

    T. Kudo, Y. Nakada, T. Matsumoto

    SPPRA 2007     222 - 227  2007

  • Bayesian Angle Information HMM with a von Mises Distribution and its Implementation using a Bayesian Monte Carlo Method

    H. Sasaki, Y. Nakada, T. Kaburagi, T. Matsumoto

    Proc. European Symposium on Time Series Prediction     29 - 38  2007

  • Online Bayesian Modeling and Prediction of Nonlinear Systems:Sequential Monte Carlo Approach

    T. Matsumoto

    Methods of Information in Medicine   46   96 - 101  2007

  • Effectiveness of Pen Pressure, Azimuth, and Altitude Features for Online Signature Verification

    Daigo Muramatsu, Takashi Matsumoto

    2nd International Conference on Biometrics, Lecture Notes in Computer Science   4642   503 - 512  2007

    DOI

  • On-line and Batch Inference for Bioinformatics Data

    T. Matsumoto

    Banff International Resarch Station Workshop on Bioinformatics, Genetics and Stochastic Computation: Bridging the Gap    2007

  • A Markov chain Monte Carlo algorithm for Bayesian dynamic signature verification

    Daigo Muramatsu, Mitsuru Kondo, Masahiro Sasaki, Satoshi Tachibana, Takashi Matsumoto

    IEEE TRANSACTIONS ON INFORMATION FORENSICS AND SECURITY   1 ( 1 ) 22 - 34  2006.03

     View Summary

    Authentication of handwritten signatures is becoming increasingly important. With a rapid increase in the number of people who access Tablet PCs and PDAs, online signature verification is one of the most promising techniques for signature verification. This paper proposes a new algorithm that performs a Monte Carlo based Bayesian scheme for online signature verification. The new algorithm consists of a learning phase and a testing phase. In the learning phase, semi-parametric models are trained using the Markov Chain Monte Carlo (MCMC) technique to draw posterior samples of the parameters involved. In the testing phase, these samples are used to evaluate the probability that a signature is genuine. The proposed algorithm achieved an EER of 1.2% against the MCYT signature corpus where random forgeries are used for learning and skilled forgeries are used for evaluation. An experimental result is also reported with skilled forgery data for learning.

    DOI

  • オンライン署名認証における経時変化適応:逐次モンテカルロ的アルゴリズム

    加藤雄大, 村松大吾, 松本隆

    2006年暗号と情報セキュリティシンポジウム予稿集   CD-ROM  2006

  • ユーザ独立型オンライン手書き署名認証モデルにおける文字種依存症に関する考察

    本郷保範, 村松大吾, 松本隆

    2006年暗号と情報セキュリティシンポジウム予稿集   CD-ROM  2006

  • 階層的ベイズ推定にもとづく顔面像認識

    松井淳, Simon Clippingdale, 鵜澤史貴, 松本隆

    画像ラボ   16 ( 3 ) 24 - 28  2006

  • 交換モンテカルロ法を用いた遺伝子制御ネットワーク推定による核内受容体ネットワークの解析

    北村悠輔, 君和田友美, 松本隆, 和田圭司

    バイオインフォマテイクス学会第10回システムバイオロジー研究会    2006

  • 変化検出機構内包モデルによるモンテカルロ自己位置推定の改良

    大家淳二, 中田洋平, 山田康平, 松本隆

    電子情報通信学会2006年総合大会(IEICE2006)   D-8-9   92  2006

  • A Sequential Monte Carlo Algorithm for Adaptation to Intersession Variability in On-line Signature Verification

    Yudai Kato, Daigo Muramatsu, Takashi Matsumoto

    IWFHR 10 International Workshop on Frontiers in Handwriting Recognition     467 - 472  2006

  • A Bayesian Change Detection for Unknown Nonlinear Systems: On-line Sequential Monte Carlo Approach

    Yohei Nakada, Takashi Matsumoto

    SCIS & ISIS 2006     1872 - 1877  2006

  • 逐次学習型顔画像認識におけるSMCサンプル数の動的制御

    松井淳, Simon Clippingdale, 松本隆

    情報科学技術レターズ 2006(FIT2006)     97 - 100  2006

  • GibbsBoostによる正面顔画像検出-事前情報を考慮するBayes的アプローチ-

    シンポジウム発表シンポジウム発表木村彰夫, 松井淳, 中田洋平, 松本隆

    第5回情報科学技術フォーラム(FIT2006)     17 - 18  2006

  • 逐次モンテカルロ動画像追跡:速度パラメータ,検索範囲の自動調整による精度向上

    中尾忠義, 松井淳, 中田洋平, 松本隆

    第5回情報科学技術フォーラム(FIT2006)     37 - 38  2006

  • ラオ・ブラックウェル逐次モンテカルロ法を用いたオンラインベイズ学習による多次元混合正規分布推定

    伊藤慶太, 中田洋平, 松本隆

    第16回 日本神経回路学会全国大会(JNNS2006)     56 - 57  2006

  • 重点サンプリング方を導入した並列型強化学習によるメタ学習

    関川昭二, 中田洋平, 松本隆

    第16回 日本神経回路学会全国大会(JNNS2006)     116 - 117  2006

  • 顔画像検出に対するベイズ敵アプローチ:逐次モンテカルロ法によるブースト学習アルゴリズム

    毛利雄介, 松井淳, 中田洋平, 松本隆

    第16回 日本神経回路学会全国大会(JNNS2006)     176 - 177  2006

  • A Sequential Monte Carlo Algorithm for Adaptation to Intersession Variability in On-line

    Yudai Kato, Daigo Muramatsu, Takashi Matsumoto

    IWFHR 10th International Workshop on Frontiers in Handwriting Recognition     467 - 472  2006

  • Signature Verification using a Monte Carlo-based Updating Algorithm Adapted to Intersession Variability

    Yudai Kato, Daigo Muramatsu, Takashi Matsumoto

    ISPACS 2006     387 - 390  2006

    DOI

  • Online Signature Verification based on User-generic Fusion Model with Markov Chain Monte Carlo Method

    Shintaro Kinoshita, Daigo Muramatsu, Takashi Matsumoto

    ISPACS 2006     391 - 394  2006

    DOI

  • A sequential Monte Carlo method for Bayesian face recognition

    Atsushi Matsui, Simon Clippingdale, Takashi Matsumoto

    STRUCTURAL, SYNTACTIC, AND STATISTICAL PATTERN RECOGNITION, PROCEEDINGS   4109   578 - 586  2006

     View Summary

    This paper proposes a Sequential Monte Carlo (SMC) learning algorithm for Bayesian probability distributions that describe model parameters in a video face recognition system based on deformable template matching. The new algorithm achieves significantly improved robustness of recognition against facial expressions and speech movements by comparison with a baseline batch MCMC (Markov Chain Monte Carlo) algorithm, at no additional computational cost. Experimental results demonstrate the effectiveness and computational efficiency of the new algorithm.

  • A Sequential Monte Carlo Algorithm for Adaptation to Intersession Variability in On-line Signature Verification

    Yudai Kato, Daigo Muramatsu, Takashi Matsumoto

    IWFHR 10 International Workshop on Frontiers in Handwriting Recognition     467 - 472  2006

  • A Bayesian Change Detection for Unknown Nonlinear Systems: On-line Sequential Monte Carlo Approach

    Yohei Nakada, Takashi Matsumoto

    SCIS & ISIS 2006     1872 - 1877  2006

  • A Sequential Monte Carlo Algorithm for Adaptation to Intersession Variability in On-line

    Yudai Kato, Daigo Muramatsu, Takashi Matsumoto

    IWFHR 10th International Workshop on Frontiers in Handwriting Recognition     467 - 472  2006

  • Signature Verification using a Monte Carlo-based Updating Algorithm Adapted to Intersession Variability

    Yudai Kato, Daigo Muramatsu, Takashi Matsumoto

    ISPACS 2006     387 - 390  2006

    DOI

  • Online Signature Verification based on User-generic Fusion Model with Markov Chain Monte Carlo Method

    Shintaro Kinoshita, Daigo Muramatsu, Takashi Matsumoto

    ISPACS 2006     391 - 394  2006

    DOI

  • Marginal likelihood change detection: Particle filter approach

    T. Matsumoto

    AIP Conference Proceedings   803 ( 1 ) 129 - 136  2005.11

     View Summary

    Given sequential data from a target system whose description is not available, this study attempts to perform online change detection by (i) using parameter/hyperparameter dynamics driven by the available data
    (ii) examining the time dependency of the marginal likelihood
    and (iii) implementing the scheme via Particle Filter. © 2005 American Institute of Physics.

    DOI

  • ギブスブースト学習アルゴリズム:逐次モンテカルロ法的アプローチ

    中田洋平, 本郷保範, 松本隆

    日本神経回路学会全国大会     72 - 73  2005

  • 自然平滑化モデルによるオンラインベイズ学習の高精度化

    中田洋平, 若原牧生, 松本隆

    日本神経回路学会全国大会     74 - 75  2005

  • モンテカルロ隠れマルコフモデルを用いたオンライン文字認識

    船田篤志, 佐々木浩人, 中田洋平, 松本隆

    日本神経回路学会全国大会     137 - 138  2005

  • 動的判断問題に対する逐次モンテカルロ法を用いたオンラインベイズ学習

    工藤智宏, 中田洋平, 松本隆

    日本神経回路学会全国大会     70 - 71  2005

  • 逐次モンテカルロ法を用いたオンラインベイズ学習による混合正規分布推定

    伊藤慶太, 中田洋平, 松本隆

    日本神経回路学会全国大会     139 - 140  2005

  • The Simultaneous Adjustment of Learning Rate and Inverse Temperature by Parallel Model in Reinforcement Learning

    Takahiro Nakamizo, Ken Achiwa, Takashi Matsumoto

    Crest Workshop    2005

  • Meta-learning and Change Detection in Reinforcement Learning with Sequential Monte Carlo

    Akio Tanaka, Ryohei Watanabe, Takashi Matsumoto

    Crest Workshop    2005

  • Sequential Monte Carloによる強化学習メタパラメータ調整

    渡辺亮平, 中田洋平, 松本隆

    電子情報通信学会 信学技法   104 ( 759 ) 107 - 112  2005

  • 動的環境下の強化学習アルゴリズム:Sequential Monte Carloとサンプル初期化

    田中昭雄, 中田洋平, 松本隆

    電子情報通信学会 信学技法   104 ( 759 ) 101 - 106  2005

  • 「逐次周辺尤度」変化検出アルゴリズム‐Seqential Monte Carlo的アプローチ‐

    松本隆

    電子情報通信学会 信学技法   104 ( 759 ) 95 - 100  2005

  • A Boosting Algorithm via Sequential Monte Carlo : GibbsBoost

    Y. Nakada, Y. Hongo, T.Matsumoto

    Proc. IASC world conference on Computational Statistics & Data Analysis    2005

  • An On-line Bayesian Learning via Natural Smoother Model

    M. Wakahara, Y. Nakada, T. Matsumoto

    Proc. IASC world conference on Computational Statistics & Data Analysis    2005

  • On-line Bayesian change detection scheme for unknown nonlinear systems via sequential Monte Carlo

    Y Nakada, T Matsumoto

    Proceedings of the International Joint Conference on Neural Networks (IJCNN), Vols 1-5     2207 - 2212  2005

     View Summary

    An attempt is made to performe on-line change detection given sequential data from an unknown nonlinear system. The algorithm sequentially estimates the probability of occurrence of a change within a Bayesian framework. The implementation is done via Sequential Monte Carlo (SMC). The proposed scheme is tested against two specific examples.

    DOI

  • Marginal Likelihood Change Detection: Particle Filter Approach

    T. Matsumoto

    AIP Conference Proceedings   803 ( 1 ) 129 - 136  2005

    DOI

  • An on-line change detection scheme for nonlinear time series data from unknown dynamical systems: A Bayesian appraoch using sequential Monte Carlo

    Y Nakada, T Matsumoto

    2005 IEEE Workshop on Machine Learning for Signal Processing (MLSP)     323 - 328  2005

     View Summary

    This paper attempts to perform on-line change detection given time series data from unknown nonlinear dynamical systems. In the algorithm, the probability of occurrence of an abrupt change is estimated within a Bayesian framework. The implementation is done via Sequential Monte Carlo (SMC). The proposed scheme is tested against two examples with nonlinear dynamical systems.

    DOI

  • Online Bayesian Modeling and Prediction of Nonlinear Systems: Sequential Monte Carlo Approach

    T. Matsumoto

    Proc. International Workshop on Biological Interpretation    2005

  • ハミルトニアンモンテカルロ法によるBayes 的顔画像認識

    松井淳, クリピングデル サイモン, 鵜沢史貴, 松本隆

    映像情報メデイア学会誌   59 ( 8 ) 1183 - 1190  2005

    DOI CiNii

  • 強化学習の並列型メタ学習:学習率の調整

    阿知波健, 渡辺亮平, 田中昭雄, 大家淳二, 松本隆

    電子情報通信学会論文誌 (D)   J88-D-1 ( 12 ) 1773 - 1784  2005

  • Modification of Intersession Variability in On-line Signature Verifier

    Y. Hongo, D. Muramatsu, T. Matsumoto

    Lecture Notes in Computer Science, Springer-Verlag   3546   455 - 463  2005

  • Natural Smoother with Hyper-paramter for On-line Bayesian Learning

    M. Wakahara, Y. Nakada, T. Matsumoto

    IBIS 2005     39 - 44  2005

  • A Boosting Algorithm via Sequential Monte Carlo : GibbsBoost

    Y. Nakada, Y. Hongo, T.Matsumoto

    Proc. IASC world conference on Computational Statistics & Data Analysis    2005

  • An On-line Bayesian Learning via Natural Smoother Model

    M. Wakahara, Y. Nakada, T. Matsumoto

    Proc. IASC world conference on Computational Statistics & Data Analysis    2005

  • AdaBoost-based on-line signature verifier

    Y Hongo, D Muramatsu, T Matsumoto

    Biometric Technology for Human Identification II   5779   373 - 380  2005

     View Summary

    Authentication of individuals is rapidly becoming an important issue. The authors previously proposed a Pen-input online signature verification algorithm. The algorithm considers a writer's signature as a trajectory of pen position, pen pressure, pen azimuth, and pen altitude that evolve over time, so that it is dynamic and biometric. Many algorithms have been proposed and reported to achieve accuracy for on-line signature verification, but setting the threshold value for these algorithms is a problem. In this paper, we introduce a user-generic model generated by AdaBoost, which resolves this problem. When user- specific models (one model for each user) are used for signature verification problems, we need to generate the models using only genuine signatures. Forged signatures are not available because imposters do not give forged signatures for training in advance. However, we can make use of another's forged signature in addition to the genuine signatures for learning by introducing a user generic model. And Adaboost is a well-known classification algorithm, making final decisions depending on the sign of the output value. Therefore, it is not necessary to set the threshold value.
    A preliminary experiment is performed on a database consisting of data from 50 individuals. This set consists of western-alphabet-based signatures provide by a European research group. In this experiment, our algorithm gives an FRR of 1.88% and an FAR of 1.60%. Since no fine-tuning was done, this preliminary result looks very promising.

    DOI

  • On-line Bayesian change detection scheme for unknown nonlinear systems via sequential Monte Carlo

    Yohei Nakada, Takashi Matsumoto

    Proceedings of the International Joint Conference on Neural Networks   4   2207 - 2212  2005

     View Summary

    An attempt is made to performe on-line change detection given sequential data from an unknown nonlinear system. The algorithm sequentially estimates the probability of occurrence of a change within a Bayesian framework. The implementation is done via Sequential Monte Carlo (SMC). The proposed scheme is tested against two specific examples. © 2005 IEEE.

    DOI

  • An on-line change detection scheme for nonlinear time series data from unknown dynamical systems: A bayesian appraoch using sequential Monte Carlo

    Yohei Nakada, Takashi Matsumoto

    2005 IEEE Workshop on Machine Learning for Signal Processing     323 - 328  2005

     View Summary

    This paper attempts to perform on-line change detection given time series data from unknown nonlinear dynamical systems. In the algorithm, the probability of occurrence of an abrupt change is estimated within a Bayesian framework. The implementation is done via Sequential Monte Carlo (SMC). The proposed scheme is tested against two examples with nonlinear dynamical systems. ©2005 IEEE.

    DOI

  • Online Bayesian Modeling and Prediction of Nonlinear Systems: Sequential Monte Carlo Approach

    T. Matsumoto

    Proc. International Workshop on Biological Interpretation    2005

  • Modification of Intersession Variability in On-line Signature Verifier

    Y. Hongo, D. Muramatsu, T. Matsumoto

    Lecture Notes in Computer Science, Springer-Verlag   3546   455 - 463  2005

  • Bayesian reconstructions and predictions of nonlinear dynamical systems via the hybrid Monte Carlo scheme

    Y Nakada, T Matsumoto, T Kurihara, K Yosui

    SIGNAL PROCESSING   85 ( 1 ) 129 - 145  2005.01

     View Summary

    Time series prediction is a rather difficult problem when the dynamics behind the data originates from a nonlinear system and its functional form is unknown. The hierarchical Bayesian scheme previously proposed by the authors has been shown to be reasonably sound for nontrivial real world applications. A great difficulty implementing the Hierarchical Bayesian scheme lies in the computation of posterior distributions as well as predictive distributions for which Quadratic Approximations have been used so far. This paper attempts to compute predictive mean and error bar for nonlinear time series prediction problems via the Hybrid Monte Carlo scheme, a particular class of Markov Chain Monte Carlo without the Quadratic Approximations. The scheme is tested against two concrete problems; Chaotic time series prediction, and Building air-conditioning Load Prediction. The prediction results are compared with those using the Quadratic Approximations which have been used in the previous works of the authors' group. The proposed scheme outperforms the Quadratic Approximations. (C) 2004 Elsevier B.V. All rights reserved.

    DOI CiNii

  • Natural Smoother with Hyper-paramter for On-line Bayesian Learning

    M. Wakahara, Y. Nakada, T. Matsumoto

    IBIS 2005     39 - 44  2005

  • Information driven parameter dynamics for on-line Bayesian learning with sequential Monte Carlo

    K Yosui, M Wakahara, Y Nakada, T Matsumoto

    ISPACS 2005: PROCEEDINGS OF THE 2005 INTERNATIONAL SYMPOSIUM ON INTELLIGENT SIGNAL PROCESSING AND COMMUNICATION SYSTEMS     377 - 380  2005

     View Summary

    A new parameter dynamics that incorporates the information available for training instead of the standard "blind" parameter dynamics is proposed for on-line Bayesian learning. A significant improvement is realized over the schemes the authors have previously proposed. The particular advantage of the currently proposed approach is the speed at which it follows abrupt changes.

  • 階層ベイズモデルとその周辺ー統計科学のフロンティア

    石黒真木夫, 松本隆, 乾敏郎, 田辺国士

    岩波書店   4  2004

  • Transmembrane Region Prediction with Hydropathy Index/Charge Two-Dimensional Trajectories of Stochastic Dynamical Systems

    Takashi Kaburagi, Daigo Muramatsu, Shinichiro Hashimoto, Masahiro Sasaki, Takashi Matsumoto

    The Second Asia Pacific Bioinformatics Conference     35 - 42  2004

  • Bayes的Markov Chain Monte Carlo顔画像認識

    鵜澤史貴, 松本隆, 松井淳, Simon Clippingdale

    電子情報通信学会 信学技法   ITS2003-91   19 - 24  2004

  • 0.18μm CMOS 2GHz Error-Correcting Encoder

    Masahiro Sasaki, Mai Nozawa, Takashi Matsumoto

    WSEAS Transactions on Circuits and Systems   3 ( 3 ) 521 - 526  2004

  • A wired CDMA interface system

    Masahiro Sasaki, Yu Ono, Takashi Matsumoto

    8th WSEAS International Conference on Circuits    2004

  • 0.18μm CMOS 6GHz Pseudo Non-overlapping Clock Generator using High-speed Dividers

    Masahiro Sasaki, Shin Yokoyama, Takashi Matsumoto

    8th WSEAS International Conference on Circuits, Athens    2004

  • A Sequential Marginal Likelihood Change Detector: Sequential Monte Carlo Approach

    T.Matsumoto

    Second Harvard Workshop on Monte Carlo Methods    2004

  • Bayesian Face Recognition using a Markov Chain Monte Carlo Method

    A. MATSUI, S. CLIPPINGDALE, F. UZAWA, T. MATSUMOTO

    International Conference on Pattern Recognition, Cambridge, U. K.    2004

    DOI

  • SMCのInformation Drivenパラメータダイナミクス

    用水邦明, 若原牧生, 松本隆

    日本神経回路学会第14回全国大会講演論文集     26 - 27  2004

  • 粒子フィルタ「逐次周辺尤度」による変化検出

    松本隆

    日本神経回路学会第14回全国大会講演論文集     28 - 29  2004

  • Transmembrane Region Prediction with Hydropathy Index/Charge Two-Dimensional Trajectories of Stochastic Dynamical Systems

    Takashi Kaburagi, Daigo Muramatsu, Shinichiro Hashimoto, Masahiro Sasaki, Takashi Matsumoto

    The Second Asia Pacific Bioinformatics Conference     35 - 42  2004

  • 0.18μm CMOS 2GHz Error-Correcting Encoder

    Masahiro Sasaki, Mai Nozawa, Takashi Matsumoto

    WSEAS Transactions on Circuits and Systems   3 ( 3 ) 521 - 526  2004

  • A wired CDMA interface system

    Masahiro Sasaki, Yu Ono, Takashi Matsumoto

    8th WSEAS International Conference on Circuits    2004

  • 0.18μm CMOS 6GHz Pseudo Non-overlapping Clock Generator using High-speed Dividers

    Masahiro Sasaki, Shin Yokoyama, Takashi Matsumoto

    8th WSEAS International Conference on Circuits, Athens    2004

  • A Sequential Marginal Likelihood Change Detector: Sequential Monte Carlo Approach

    T.Matsumoto

    Second Harvard Workshop on Monte Carlo Methods    2004

  • Bayesian face recognition using a Markov chain Monte Carlo method

    Atsushi Matsui, Simon Clippingdale, Fumiki Uzawa, Takashi Matsumoto

    Proceedings - International Conference on Pattern Recognition   3   918 - 921  2004

     View Summary

    A new algorithm is proposed for face recognition by a Bayesian framework. Posterior distributions are computed by Markov chain Monte Carlo (MCMC). Face features used in the paper are those used in our previous work[1][2] based on the Elastic Graph Matching method. While our previous method attempts to optimize facial feature point positions so as to maximize a similarity function between each model and face region in the input sequence, the proposed approach evaluates posterior distributions of models conditioned on the input sequence. Experimental results show a rather dramatic improvement in robustness. The proposed algorithm eliminates almost all identification errors on sequences showing individuals talking, and reduces identification errors by more than 90% on sequences showing individuals smiling although such data was not used in training.

    DOI

  • The reduction of memory and the improvement of recognition rate for HMM on-line handwriting recognition

    A Funada, D Muramatsu, T Matsumoto

    NINTH INTERNATIONAL WORKSHOP ON FRONTIERS IN HANDWRITING RECOGNITION, PROCEEDINGS     383 - 388  2004

     View Summary

    The purpose of this project is two fold. The first purpose is to reduce the memory size of our previous handwriting recognition algorithm based on an HMM using Self-Organizing Map (SOM) density tying. The second is to improve recognition capability by incorporating additional information. SOM density tying reduced the dictionary size to 1/7 of the original size, with a recognition rate of 90.45%, only slightly less than the original recognition rate of 91.51%. Our additional feature increased recognition capability to 91.34%.

    DOI

  • Linearity performance comparison of cascode current source and single-device current source IDPs; analyses, simulations and measurements

    K Hadidi, M Morimoto, K Futami, T Oue, M Ito, M Sasaki, A Khoei, T Matsumoto

    INTERNATIONAL JOURNAL OF ELECTRONICS   90 ( 5 ) 341 - 353  2003.05

     View Summary

    The input differential pair (IDP) is usually a major source of nonlinear distortion in any op-amp. This is especially true if the input signal has a large common-mode component, as is the case when an op-amp functions as a unity-gain buffer or as part of a single-ended sample-hold (S/H) circuit. In this paper, we analyse the distortion of the commonly used cascode current source IDP structure and explain the sources of its nonlinear behaviour. Next, a special design technique is proposed which enhances the linearity of IDPs. The circuit uses a single device current source that has the same channel length while its width is double those of IDP devices. Theoretical analysis, as well as simulation and experimental results, is given to confirm the improved linearity of a unity gain buffer. Simulations predict improvements up to 20 dB. 15 dB total harmonic distortion (THD) reduction was also achieved for a 15 MHz input signal based on measurement of a test chip. The method is valuable as power supply voltages shrink, and the design offers extra voltage headroom at input.

    DOI CiNii

  • A low power matched filter for DS-CDMA based on analog signal processing

    M Sasaki, T Sakai, T Matsumoto

    IEICE TRANSACTIONS ON FUNDAMENTALS OF ELECTRONICS COMMUNICATIONS AND COMPUTER SCIENCES   E86A ( 4 ) 752 - 757  2003.04

     View Summary

    This paper proposes a low power consumption Analog Matched Filter (AMF) that utilizes capacitor multiply-and-accumulate operations. A high-speed, high-precision Analog-to-Digital (A/D) converter is unnecessary because the proposed circuit directly samples received analog signals. A code-shifting MF structure is used to prevent errors from accumulating. A 15-tap AMF circuit was fabricated using 0.35 mum CMOS technology. Power consumption for the 128-tap circuit is estimated to be 22.3 mW at 25 MHz and 3.3 V, and the area is estimated to be 0.33 mm(2). The proposed circuit will thus be a useful LSI for mobile terminals.

  • low-discrepancy列を用いた Markov Chain Monte Carlo

    和田健作, 松本隆

    電子情報通信学会総合大会   D-2-5   11  2003

  • ペン位置軌跡情報を用いたHMM On-line 署名照合

    村松大吾, 松本隆

    電子情報通信学会総合大会   D-12-24   185  2003

  • Bayes的Sequential Monte Carlo学習における Rao-Blackwellisationの効果

    相馬貴也, 用水邦明, 松本隆

    電子情報通信学会総合大会   D-2-4   10  2003

  • 有線CDMA インターフェースシステムの検討

    小野祐, 佐々木昌浩, 松本隆

    電子情報通信学会総合大会   C-12-43   112  2003

  • 遺伝的アルゴリズムを用いた強化学習メタパラメータの学習

    窪優司, 渡邉亮平, 松本隆

    電子情報通信学会総合大会   D-2-18   24  2003

  • Bayesian MCMC オンライン署名照合

    近藤充, 村松大吾, 佐々木昌浩, 松本隆

    電子情報通信学会総合大会   D-12-25   186  2003

  • 強化学習における価値関数の分散最小化によるメタ学習

    阿知波健, 松本隆

    電子情報通信学会総合大会   D-8-29   123  2003

  • 2.5V CMOS Fully Differential Low Power High Linearity Analog Line-Driver

    Khayrollah Hadidi, 大島宗之, 佐々木昌浩, 松本隆

    電子情報通信学会総合大会   C-12-42   111  2003

  • CMOS スイッチトキャパシタDC-DCコンバータ

    Kayrollah Hadidi, 釣井雄介, 佐々木昌浩, 松本隆

    電子情報通信学会総合大会   C-12-33   102  2003

  • 2.5V CMOS Fully Differential Low Power High Linearity Analog Line-Driver

    Khayrollah Hadidi, 大島宗之, 佐々木昌浩, 松本隆

    第16回回路とシステム(軽井沢)ワークショップ論文集     67 - 72  2003

  • An HMM On-line Signature Verification with Pen Position Trajectories

    D.Muramatsu, T.Matsumoto

    The 2003 International Conference on Artificial Intelligence (IC-AI 2003)     299 - 303  2003

  • Bayesian MCMC for biometric person authentication incorporating on-line signature trajectories

    Mitsuru Kondo, Daigo Muramatsu, Masahiro Sasaki, Takashi Matsumoto

    Proceedings of the IASTED International Conference on Signal Processing, Pattern Reconition, and Applications     269 - 273  2003

     View Summary

    Authentication of individuals is rapidly becoming an important issue. The authors have previously proposed a pen-input online signature verification algorithm. The algorithm considers writer's signature as a trajectory of pen-position, pen-pressure and pen-inclination which evolves over time, so that it is dynamic and biometric. In our previous work, genuine signatures were separated from forgery signatures in a linear manner. This paper proposes a new algorithm which performs nonlinear separation using Bayesian MCMC (Markov Chain Monte Carlo). A preliminary experiment is performed on a database consisting of 1825 genuine signatures and 4117 skilled forgery signatures from fourteen individuals. FRR 0.81% and FAR 0.87% are achieved. Since no fine tuning was done, this preliminary result looks very promising.

  • A Bayesian MCMC On-line Signature Verification

    Mitsuru Kondo, Daigo Muramatsu, Masahiro Sasaki, Takashi Matsumoto

    4th International Conference on Audio-and Video-Based Biometric Person Authentication     540 - 548  2003

  • ペン変位情報のみを用いたHMMオンライン署名照合手法の検討

    村松大吾, 松本隆

    ユビキタスネットワーク社会におけるバイオメトリクスセキュリティ研究会、第一回研究発表会予稿集     1 - 5  2003

  • ベイズ的モンテカルロ手法を用いたオンライン手書き署名照合

    近藤充, 村松大吾, 佐々木昌浩, 松本隆

    ユビキタスネットワーク社会におけるバイオメトリクスセキュリティ研究会、第一回研究発表会予稿集     7 - 12  2003

  • Internet packet transmission delay prediction with sequential Monte Carlo

    Takayuki Kurihara, Ryosuke Tsuboi, Takashi Matsumoto

    Practical Bayesian Statistics 5     59  2003

  • Bayesian Reconstructions and Predictions of Nonlinear Dynamical Systems by Rao-Blackwellised Sequential Monte Carlo

    Takaya Souma, Kuniaki Yosui, Takashi Matsumoto

    Practical Bayesian Statistics 5    2003

  • A Bayesian MCMC On-Line Algorithm for Signature Verification

    Mitsuru Kondo, Daigo Muramatsu, Masahiro Sasaki, Takashi Matsumoto

    Practical Bayesian Statistics 5     34  2003

  • ベイズ的非線形ダイナミカルシステムの再構成と予測:Hybrid Monte Carlo

    中田洋平, 栗原貴之, 用水邦明, 和田健作, 松本隆

    電子情報通信学会論文誌   J86-D-II ( 8 ) 1143 - 1155  2003

  • Rao-Blackwellised Sequential Monte Carloによる非線形ダイナミカルシステムの再構成と予測

    相馬貴也, 松本隆

    2003年電気学会 電子・情報・システム部門大会    2003

  • An HMM on-line signature verifier incorporating signature trajectories

    Daigo Muramatsu, Takashi Matsumoto

    Proceedings of the International Conference on Document Analysis and Recognition, ICDAR   2003-   438 - 442  2003

     View Summary

    Authentication of individuals is rapidly becoming an important issue. On-line signature verification is one of the methods that use biometric features. This paper proposes a new HMM algorithm is for on-line signature verification. After preprocessing, input signature is discretized in a polar coordinate system. This particular discretization leads to a simple procedure for assigning initial state and state transition probabilities. This paper utilizes only pen position trajectories, no other information is used which makes the algorithm simple and fast. A preliminary experiment shows that the proposed algorithm appears to be promising.

    DOI

  • ダイナミカルシステムの階層ベイズ的最小埋め込み次元推定

    杉淳二郎, 栗原貴之, 松本隆

    情報処理学会論文誌   44 ( 12 ) 3098 - 3111  2003

  • カオスに関する最近の研究動向

    カオスとその周辺問題調査専門委員会

    電気学会技術報告   929   23  2003

  • Reconstructions and predictions of nonlinear dynamical systems by Rao-Blackwellised sequential Monte Carlo

    T Soma, K Yosui, T Matsumoto

    2003 IEEE INTERNATIONAL CONFERENCE ON ACOUSTICS, SPEECH, AND SIGNAL PROCESSING, VOL II, PROCEEDINGS   II   777 - 780  2003

     View Summary

    Sequential Monte Carlo (SMC) is a powerful sampling based inference/learning algorithm for Bayesian scheme. The purpose of this paper is two fold. It first attempts to reconstruct and predict nonlinear dynamical systems from one dimensional data which arrives in a sequential manner instead of batch manner. Second purpose is to test the performance of the Rao-Blackwellisation in reconstructing and predicting nonlinear dynamical systems. We demonstrate that Rao-Blackwellised Sequential Monte Carlo (RBSMC) on a chaotic time series prediction problem outperforms generic SMC.

  • An HMM On-line Signature Verification with Pen Position Trajectories

    D.Muramatsu, T.Matsumoto

    The 2003 International Conference on Artificial Intelligence (IC-AI 2003)     299 - 303  2003

  • Bayesian MCMC for Biometric Person Authentication Incorporating On-Line Signature Trajectories

    Mitsuru Kondo, Daigo Muramatsu, Masahiro Sasaki, Takashi Matsumoto

    IASTED International Conference on Signal Processing, Pattern Recognition and Applications (SPPRA 2003)     269 - 273  2003

     View Summary

    Authentication of individuals is rapidly becoming an important issue. The authors have previously proposed a pen-input online signature verification algorithm. The algorithm considers writer's signature as a trajectory of pen-position, pen-pressure and pen-inclination which evolves over time, so that it is dynamic and biometric. In our previous work, genuine signatures were separated from forgery signatures in a linear manner. This paper proposes a new algorithm which performs nonlinear separation using Bayesian MCMC (Markov Chain Monte Carlo). A preliminary experiment is performed on a database consisting of 1825 genuine signatures and 4117 skilled forgery signatures from fourteen individuals. FRR 0.81% and FAR 0.87% are achieved. Since no fine tuning was done, this preliminary result looks very promising.

  • A Bayesian MCMC On-line Signature Verification

    Mitsuru Kondo, Daigo Muramatsu, Masahiro Sasaki, Takashi Matsumoto

    4th International Conference on Audio-and Video-Based Biometric Person Authentication     540 - 548  2003

  • An HMM on-line signature verification algorithm

    D Muramatsu, T Matsumoto

    AUDIO-AND VIDEO-BASED BIOMETRIC PERSON AUTHENTICATION, PROCEEDINGS   2688   233 - 241  2003

     View Summary

    Authentication of individuals is rapidly becoming an important issue. On-line signature verification is one of the methods that use biometric features of individuals. This paper proposes a new HMM algorithm for on-line signature verification incorporating signature trajectories. The algorithm utilizes only pen position trajectories. No other information is used which makes the algorithm simple and fast. A Preliminary experiment was performed and the intersection of FAR and FRR was 2.78%.

  • Internet packet transmission delay prediction with sequential Monte Carlo

    Takayuki Kurihara, Ryosuke Tsuboi, Takashi Matsumoto

    Practical Bayesian Statistics 5     59  2003

  • Bayesian Reconstructions and Predictions of Nonlinear Dynamical Systems by Rao-Blackwellised Sequential Monte Carlo

    Takaya Souma, Kuniaki Yosui, Takashi Matsumoto

    Practical Bayesian Statistics 5    2003

  • Nonlinear separation of signature trajectories for on-line personal authentication

    M Kondo, D Muramatsu, M Sasaki, T Matsumoto

    2003 INTERNATIONAL CONFERENCE ON MULTIMEDIA AND EXPO, VOL III, PROCEEDINGS     89 - 92  2003

     View Summary

    Authentication of individuals is rapidly becoming an important issue. This paper proposes a new nonlinear algorithm for pen-input on-line signature verification incorporating pen-position, pen-pressure and pen-inclinations trajectories.
    A preliminary experiment is performed on a database consisting of 1849 genuine signatures and 3174 skilled(dagger) forgery signatures from fourteen individuals. Since no fine tuning was done, this preliminary result looks very promising.

  • A Bayesian MCMC On-Line Algorithm for Signature Verification

    Mitsuru Kondo, Daigo Muramatsu, Masahiro Sasaki, Takashi Matsumoto

    Practical Bayesian Statistics 5     34  2003

  • An HMM on-line signature verifier incorporating signature trajectories

    D Muramatsu, T Matsumoto

    SEVENTH INTERNATIONAL CONFERENCE ON DOCUMENT ANALYSIS AND RECOGNITION, VOLS I AND II, PROCEEDINGS     438 - 442  2003

     View Summary

    Authentication of individuals is rapidly becoming an important issue. On-line signature verification is one of the methods that use biometric features. This paper proposes a new HMM algorithm is for on-line signature verification. After preprocessing, input signature is discretized in a polar coordinate system. This particular discretization leads to a simple procedure for assigning initial state and state transition probabilities.
    This paper utilizes only pen position trajectories, no other information is used which makes the algorithm simple and fast. A preliminary experiment shows that the proposed algorithm appears to be promising.

    DOI

  • A highly linear fully differential low power CMOS line driver

    K Hadidi, H Oshima, M Sasaki, T Matsumoto

    ESSCIRC 2003: PROCEEDINGS OF THE 29TH EUROPEAN SOLID-STATE CIRCUITS CONFERENCE     541 - 544  2003

     View Summary

    fA differential wide-band line driver for transformer-coupled cables is presented. Implemented in a 0.18 mum CMOS process, its -3dB bandwidth is 130MHz. The new line driver consumes 60mW of static power from a 2.5V supply. Its total power driving 75 Omega load is 130mW. 130mW from a 2.5V supply. The device drives a 75 Omega load, while achieves a -51.0dB THD with a 2.5V(p-p) 5MHz signal on load. The THD is -48.9dB for a 2.5V(p-p) 10MHz output signal. The area for the device is less than 0.1mm(2).

    DOI

  • Inferring transmembrane region counts with hydropathy index/charge two dimensional trajectories of stochastic dynamical systems

    D Muramatsu, S Hashimoto, T Tsunashima, T Kaburagi, M Sasaki, T Matsumoto

    2003 IEEE XIII WORKSHOP ON NEURAL NETWORKS FOR SIGNAL PROCESSING - NNSP'03     101 - 110  2003

     View Summary

    A new algorithm is proposed for inferring the number of transmembrane regions of transmembrane proteins from two dimensional vector trajectories consisting of hydropathy index and charge of amino acids by stochastic dynamical system models. The prediction accuracy of a preliminary experiment is 94%. Since no fine-tuning is done, this appears encouraging.

    DOI

  • From data to nonlinear dynamics: Time series prediction via hierarchical Bayes approach

    T Matsumoto, H Hamagishi, J Sugi, M Saito, Y Chonan

    ELECTRONICS AND COMMUNICATIONS IN JAPAN PART II-ELECTRONICS   85 ( 10 ) 71 - 83  2002.10

     View Summary

    In the framework of the hierarchical Bayesian approach, a method for nonlinear time series prediction using neural nets is proposed, and its particular applications are shown by examples. (C) 2002 Wiley Periodicals, Inc.

    DOI

  • Reconstructions and Predictions of Nonlinear Dynamical Systems Weighted by Model Marginal Likelihoods - A Hierarchical Bayesian Approach

    SAITO Motoki, ENOMOTO Tsuyoshi, MATSUMOTO Takashi

    Transactions of the Institute of Systems, Control and Information Engineers   15 ( 1 ) 10 - 16  2002

     View Summary

    A hierarchical Bayesian approach is formulated for nonlinear time series prediction problems with neural nets. The proposed scheme consists of several steps : <BR>(i) Formulae for posterior distributions of parameters, hyper parameters as well as models via Bayes formula.<BR>(ii) Derivation of predictive distributions of future values taking into account model marginal likelihoods.<BR>(iii) Using several drastic approximations for computing predictive mean of time series incorporating model marginal likelihoods.<BR>The proposed scheme is tested against two examples; (A) Time series data generated by noisy chaotic dynamical system, and (B) Building air-conditioning load prediction problem. The proposed scheme outperforms the algorithm previously used by the authors.

    DOI CiNii

  • 非線形ダイナミカルシステムの再構成と予測 : ベイズ的アプローチ

    松本 隆

    電子情報通信学会NC研究会    2002

  • 2次元連続区分線形同相写像におけるカオス的アトラクタの存在証明

    蟹江幸司, 徳永隆冶, 松本 隆

    電子情報通信学会論文誌   J84-A ( 1 ) 66 - 75  2002

  • Rao-Blackwellised Sequential Monte Carloによるニューラルネットの学習

    和田健作, 用水邦明, 松本隆

    電子情報通信学会総合大会   D-2-3   12  2002

  • Sequential Monte CarloによるHyperparameterのOn-Line学習

    相馬貴也, 栗原貴之, 松本隆

    電子情報通信学会総合大会   D-2-6   15  2002

  • 小量学習データセットでのHMMオンライン文字認識アルゴリズム

    石井友忠, 赤松謙, 持田誠一郎, 佐々木昌浩, 松本隆

    電子情報通信学会総合大会   D-12-69   245  2002

  • HMMオンライン手書き文字認識におけるパラメータスムージング効果

    石井友忠, 赤松謙, 佐々木昌浩, 松本隆

    情報処理学会第64回全国大会論文集   43 ( 2 ) 125 - 126  2002

  • 低解像度のタブレットにおける署名照合

    森田光, 坂本大輔, 小宮義光, 佐々木昌浩, 松本隆

    電子情報通信学会総合大会   D-12-63   239  2002

  • 電子回路のカオス同期とマスキング

    松本 隆

    電子情報通信学会論文誌   J85-C ( 3 ) 125 - 133  2002

  • 相対ペントラジェクトリーによるオンライン署名照合

    近藤充, 小宮義光, 坂本大輔, 森田光, 佐々木昌浩, 松本 隆

    電子情報通信学会総合大会   A-7-11   221  2002

  • S/H回路を用いたCMOSマッチトフィルタの設計

    河津大志, 坂井丈泰, 佐々木昌浩, 松本隆

    電子情報通信学会総合大会   C-12-32   114  2002

  • 2次元DCTインテリジェントイメージセンサの試作

    伊藤和則, 釣井雄介, 佐々木昌浩, 坂井丈泰, 松本 隆

    電子情報通信学会総合大会   C-12-48   130  2002

  • A Constant Bandwidth CMOS VGA ckt.

    Khayrollah Hadidi, 伊東充吉, 大島宗之, 佐々木昌浩, 松本 隆

    電子情報通信学会総合大会   C-12-43   125  2002

  • 低消費電力アナログマッチトフィルタ

    佐々木昌浩, 坂井丈泰, 松本 隆

    第15回回路とシステム軽井沢ワークショップ論文集     37 - 40  2002

  • 相対ペントラジェクトリーによるオンライン署名照合

    近藤 充, 小宮義光, 坂本大輔, 森田 光, 佐々木昌浩, 松本 隆

    第15回回路とシステム軽井沢ワークショップ論文集     517 - 522  2002

  • A Constant Bandwidth CMOS VGA ckt.

    Khayrollah Hadidi, 伊東充吉, 大島宗之, 佐々木昌浩, 松本 隆

    第15回回路とシステム軽井沢ワークショップ論文集     363 - 367  2002

  • Bayesian Nonlinear Time Series Predictions via Hybrid Monte Carlo

    Y. Nakada, T. Kurihara, K. Yosui, K. Wada, T. Matsumoto

    7th Valencia International Meetings on Bayesian Statistics    2002

  • Learning Hyperparameters via On-Line Bayesian Scheme

    T. Kurihara, Y. Nakada, K. Yosui, T. Matsumoto

    7th Valencia International Meetings on Bayesian Statistics    2002

  • On line Signature Verifieer Incorporating Position, Pressure, Inclination and Velocity Trajectories

    Mitsuru Kondo, Daisuke Sakamoto, Masahiro Sasaki, Takashi Matsumoto

    The 2002 International Conference on Security and Management(SAM'02)    2002

  • Rao-Blackwellised Sequential Monte Carlo のニューラルネットワークへの適用

    用水邦明, 和田健作, 松本隆

    日本神経回路学会第12回全国大会(JNNS2002とっとり)     33 - 33  2002

  • Rao-Blackwellised Sequential Monte Carloによる非線型ダイナミカルシステムの再構成と予測

    相馬貴也, 用水邦明, 松本 隆

    第5回情報論的学習理論ワークショップ(IBIS2002)予稿集     238 - 243  2002

  • Sequential Monte Carloを用いたOn-line学習におけるパラメータのグループ化の効果

    栗原貴之, 相馬貴也, 松本隆

    日本神経回路学会第12回全国大会(JNNS2002とっとり)     141 - 144  2002

  • Reconstructions and Predictions of Nonlinear Dynamical Systems via Rao-Blackwellised Sequential Monte Carlo

    Takaya Souma, Kuniaki Yosui, Takashi Matsumoto

    First Cape Cod Workshop on Monte Carlo Methods    2002

  • CMOS Analog Matched Filter Using Sample-and-Hold Circuit

    KAWATSU H.

    IEEJ International Analog VLSI Workshop , September, 2002     160 - 164  2002

    CiNii

  • 階層ベイズ的空調機熱負荷予測-データからダイナミカルシステムへ-

    松本隆, 中島芳徳, 斎藤幹貴, 杉淳二郎, 浜岸弘明

    日本気象学会 気象研究ノート203号   上   159 - 171  2002

  • Chaotic Time Series Prediction via MCMC

    Y. Nakada, K. Yosui, K. Wada, T. Kurihara, T. Matsumoto

    1st International Conference on Soft Computing and Intelligent Systems and 3rd International Symposium on Advanced Intelligent Systems (SCIS & ISIS 2002)    2002

  • カオスと時系列-情報数理シリーズB-7-

    松本隆, 徳永隆治, 宮野尚哉, 徳田功

    培風館    2002

  • A New Online Signature Verification Algorithm Incorporating Pen Velocity Trajectories

    Mitsuru Kondo, Daisuke Sakamoto, Masahiro Sasaki, Takashi Matsumoto

    2002 International Symposium on Intelligent Signal Processing and Communication Systems (ISPACS 2002)    2002

  • Dynamic Biometric Person Authentication Using Pen Signature Trajectories

    Daisuke Sakamoto, Mitsuru Kondo, Hikaru Morita, Daigo Muramatsu, Masahiro Sasaki, Takashi Matsumoto

    9th International Conference on Neural Information Processing (ICONIP’02)    2002

    DOI

  • Sequential Monte Carlo learning with hyperparameter adjustments

    K Wada, K Yosui, Y Nakada, T Matsumoto

    PROCEEDING OF THE 2002 INTERNATIONAL JOINT CONFERENCE ON NEURAL NETWORKS, VOLS 1-3     274 - 279  2002

     View Summary

    Sequential Monte Carlo scheme is proposed for on-line Bayesian learning. The proposed scheme adjusts not only parameters for data fitting but adjust hyperparameters online so that the scheme attempts to avoid over fitting in an adaptive manner. The scheme is tested against simple examples and is shown to be functional.

  • Bayesian Nonlinear Time Series Predictions via Hybrid Monte Carlo

    Y. Nakada, T. Kurihara, K. Yosui, K. Wada, T. Matsumoto

    7th Valencia International Meetings on Bayesian Statistics    2002

  • Learning Hyperparameters via On-Line Bayesian Scheme

    T. Kurihara, Y. Nakada, K. Yosui, T. Matsumoto

    7th Valencia International Meetings on Bayesian Statistics    2002

  • On line Signature Verifieer Incorporating Position, Pressure, Inclination and Velocity Trajectories

    Mitsuru Kondo, Daisuke Sakamoto, Masahiro Sasaki, Takashi Matsumoto

    The 2002 International Conference on Security and Management(SAM'02)    2002

  • Reconstructions and Predictions of Nonlinear Dynamical Systems via Rao-Blackwellised Sequential Monte Carlo

    Takaya Souma, Kuniaki Yosui, Takashi Matsumoto

    First Cape Cod Workshop on Monte Carlo Methods    2002

  • Bayesian on-line learning: A sequential Monte Carlo with Rao-Blackwellization

    K Yosui, T Kurihara, K Wada, T Souma, T Matsumoto

    NEURAL NETWORKS FOR SIGNAL PROCESSING XII, PROCEEDINGS     99 - 108  2002

     View Summary

    This paper proposes a Rao-Blackwellised Sequential Monte Carlo (RBSMC) scheme for on-line learning with feed forward neural nets. The proposed algorithm is tested against an example and the performance is compared with those of the conventional Sequential Monte Carlo as well as the Extended Kalman Filter (EKF). The proposed scheme outperforms those conventional algorithms.

    DOI

  • CMOS Analog Matched Filter Using Sample-and-Hold Circuit

    H. Kawatsu, M. Sasaki, T. Sakai, T. Matsumoto

    2002 IEEJ International Analog VLSI Workshop     160 - 164  2002

  • Chaotic Time Series Prediction via MCMC

    Y. Nakada, K. Yosui, K. Wada, T. Kurihara, T. Matsumoto

    1st International Conference on Soft Computing and Intelligent Systems and 3rd International Symposium on Advanced Intelligent Systems (SCIS & ISIS 2002)    2002

  • A New Online Signature Verification Algorithm Incorporating Pen Velocity Trajectories

    Mitsuru Kondo, Daisuke Sakamoto, Masahiro Sasaki, Takashi Matsumoto

    2002 International Symposium on Intelligent Signal Processing and Communication Systems (ISPACS 2002)    2002

  • Dynamic Biometric Person Authentication Using Pen Signature Trajectories

    Daisuke Sakamoto, Mitsuru Kondo, Hikaru Morita, Daigo Muramatsu, Masahiro Sasaki, Takashi Matsumoto

    9th International Conference on Neural Information Processing (ICONIP’02)    2002

    DOI

  • Reconstructions and predictions of nonlinear dynamical systems: A hierarchical Bayesian approach

    T Matsumoto, Y Nakajima, M Saito, J Sugi, H Hamagishi

    IEEE TRANSACTIONS ON SIGNAL PROCESSING   49 ( 9 ) 2138 - 2155  2001.09

     View Summary

    An attempt is made to reconstruct model nonlinear dynamical systems from scalar time series data via a hierarchical Bayesian framework. Reconstruction is performed by fitting given training data with a parameterized family of functions without overfitting. The reconstructed model dynamical systems are compared with respect to (approximated) model marginal likelihood, which is a natural Bayesian information criterion. The best model is selected with respect to this criterion and is utilized to make predictions. The results are applied to two problems: i) Chaotic time series prediction and ii) building air-conditioning load prediction. The former is a very good class of problems for checking abilities of prediction algorithms for at least two reasons. First, since no linear dynamical systems can admit chaotic behavior, an algorithm must capture nonlinearities behind the time series. Second, chaotic dynamical systems are sensitive to initial conditions. More precisely, the error grows exponentially with respect to time so that crispness of capturing nonlinearities is also important. Experimental results appear to indicate that the proposed scheme can capture difficult nonlinearities behind chaotic time series data. The latter class of problems (air conditioning load prediction) is motivated by a great amount of demand for reducing CO2 emissions associated with electric power generation. The authors won a prediction competition using the proposed algorithm; therefore, it appears to be reasonably sound.

    DOI CiNii

  • A pen input on-line signature verifier integrating position, pressure and inclination trajectories

    Y Komiya, T Ohishi, T Matsumoto

    IEICE TRANSACTIONS ON INFORMATION AND SYSTEMS   E84D ( 7 ) 833 - 838  2001.07

     View Summary

    Personal identity verification has a great variety of applications including access to computer terminals, buildings, credit card verification as well as EC. Algorithm for personal identity verification call be roughly classified into four categories depending on static/dynamic and biometric/physical or knowledge based. Finger prints, iris, retina, DNA, face, blood vessels, for instance, are static and biometric. Algorithms which are biometric and dynamic include lip movements, body movements and on-line signatures. Schemes which use passwords are static and knowledge based, whereas methods using magnetic cards and IC cards are physical. Each scheme naturally has its own advantages and disadvantages. A new algorithm is proposed for pen-input on-line signature verification incorporating pen-position, pen-pressure and pen-inclinations trajectories. A preliminary experiment is performed on a data base consisting uf 293 genuine writings and 540 forgery writings, from 8 individuals. Average correct verification late was 97.6% whereas average forgery refection rate was 98.7%. Since no fine tuning was done, this preliminary result looks very promising.

  • 複雑系の事典―適応複雑系のキーワード150-

    松本 隆

    朝倉書店    2001

  • Chaotic Masking における情報信号増幅作用について

    永井立紀, 浜野英知, 吉田倫己, 相田能之, 畠澤泰成, 松本 隆

    電子情報通信学会論文誌A   J84-A ( 2 ) 164 - 170  2001

  • Bayesian Nonlinear Time Series Prediction via MCMC

    Y.Nakada, T.Kurihara, N.Takahashi, K.Yosui, T.Matsumoto

    2001 IEEE-EURASIP Workshop on Nonlinear Signal and Image Processing    2001

  • 顔のGabor ウェーブレット特徴量の推定

    山根智文, 松本隆, サイモンクリピングデル, 伊藤崇之

    電子情報通信学会技術研究報告   100 ( 701 ) 119 - 124  2001

  • On-line Bayesian Meta-Learning: A Preliminary Report

    Y.Nakada, T.Kurihara, K.Yosui, T.Matsumoto

    CREST Workshop on Metalearning and Neuromodulation    2001

  • ペン位置・筆圧・傾き情報を用いた書名照合アルゴリズムPPI

    坂本大輔, 大石 哲, 小宮義光, 松本 隆

    第14回回路とシステムワークショップ論文集     513 - 518  2001

  • On-line Bayes メタ学習 : パラメータ、ハイパラメータの適応的調整

    栗原貴之, 用水邦明, 松本 隆

    クレスト「脳を創る」シンポジウム    2001

  • Markov Chain Monte Carloを用いた階層Bayes的非線形時系列予測

    用水邦明, 松本 隆

    複雑系シンポジウム    2001

  • Bayes 的 Online 学習 : Importance Resampling

    中田洋平, 栗原貴之, 用水邦明, 松本 隆

    2001年情報論的学習理論ワークショップ(IBIS2001)     189 - 193  2001

  • Sequential Monte Carloによる逐次学習

    栗原貴之, 用水邦明, 松本 隆

    複雑系シンポジウム    2001

  • On-line Monte Carlo 学習

    栗原貴之, 中田洋平, 用水邦明, 松本 隆

    京都大学数理解析研究所考究録    2001

  • ベイズ的オンライン学習におけるハイパラメータの学習とその効果

    用水邦明, 栗原貴之, 松本 隆

    日本神経回路学会第11回全国大会講演論文集(JNNS2001)     75 - 76  2001

  • Baysian On-line Learning: A Sequential Monte Carlo with Importance Resampling

    T.Kurihara, Y.Nakada, K.Yosui, T.Matsumoto

    NNSP2001     163 - 193  2001

  • カオスを用いた秘話通信における情報信号の増幅作用の解析

    畠澤泰成, 松本 隆

    情報処理学会第63回全国大会講演論文集   2   101 - 102  2001

  • ペン位置・筆圧・傾き・時間情報を用いたオンライン書名照合

    坂本大輔, 森田 光, 小宮義光, 松本 隆

    情報処理学会第63回全国大会講演論文集   3   545 - 546  2001

  • A 500MS/Sec-54dB THD Open-Loop CMOS Sample-and-Hold Stage

    佐々木昌浩, 村松大吾, 松本 隆, Hadidi Khayrollah

    電子情報通信学会エレクトロニクスソサエティ大会講演論文集   2   83  2001

  • Sequential Monte Carlo : Hyperparameter の学習

    栗原貴之, 用水邦明, 松本 隆

    日本神経回路学会第11回全国大会講演論文集(JNNS2001)     77 - 78  2001

  • オフセット補償機能を持つ差動増幅器の試作

    伊藤和則, 佐々木昌浩, 坂井丈泰, 松本 隆

    電子情報通信学会エレクトロニクスソサエティ大会講演論文集   2   94  2001

  • A 'Variable Capacitance' Circuit for Tuning Integrated Filters and Oscillators

    Kh. Hadidi, 伊東充吉, 佐々木昌浩, 松本 隆

    電子情報通信学会エレクトロニクスソサエティ大会講演論文集   2   88  2001

  • Bayesian Bi-gram Schemeを用いたHMMによるオンライン手書き文字認識アルゴリズム

    長谷川智希, 石井友忠, 赤松 謙, 松本 隆

    ヒューマンインターフェースシンポジウム2001論文集     311 - 314  2001

  • マハラノビス距離によるオンライン署名照合アルゴリズムPPI

    大石 哲, 坂本大輔, 森田 光, 小宮義光, 松本 隆

    ヒューマンインターフェースシンポジウム2001論文集     59 - 62  2001

  • A Sequential Monte Carlo Bayesian On-line Learning

    T. Kurihara, Y. Nakada, K. Yosui, T. Matsumoto

    The 9th International Symposium on Intelligent Signal Processing and Communications Systems     381 - 385  2001

  • A Pen-input On-line Signature Verifier

    H. Morita, D. Sakamoto, T. Ohishi, T. Matsumoto

    The 9th International Symposium on Intelligent Signal Processing and Communications Systems     376 - 380  2001

  • Automatical Learning of Hyperparameters via Sequential Monte Carlo

    K. Yosui, T. Kurihara, T. Matsumoto

    International Symposium New Trends in Optimization and Computational Algorithm     121 - 122  2001

  • Bayesian Nonlinear Time Series Prediction via MCMC

    Y.Nakada, T.Kurihara, N.Takahashi, K.Yosui, T.Matsumoto

    2001 IEEE-EURASIP Workshop on Nonlinear Signal and Image Processing    2001

  • Topological horseshoe in the R-L-diode circuit

    N Takeuchi, T Nagai, T Matsumoto

    ELECTRONICS AND COMMUNICATIONS IN JAPAN PART III-FUNDAMENTAL ELECTRONIC SCIENCE   84 ( 3 ) 91 - 100  2001

     View Summary

    In spite of its simplicity, numerous bifurcation phenomena are observed in the R-L-diode circuit. There are also reports on Pecora-Carroll chaos synchronization and chaotic masking. This paper shows that the topological horseshoe is built in this circuit, which guarantees the complex behavior. The intention of the investigation is to establish the theoretical aspects of the circuit. (C) 2000 Scripta Technica.

    DOI

  • RAV (reparameterized angle variations) algorithm for online handwriting recognition

    M. Kobayashi, S. Masaki, O. Miyamoto, Y. Nakagawa, Y. Komiya, T. Matsumoto

    International Journal on Document Analysis and Recognition   3 ( 3 ) 181 - 191  2001

     View Summary

    A new algorithm RAV (reparameterized angle variations) is proposed which makes explicit use of trajectory information where the time evolution of the pen coordinates plays a crucial role. The algorithm is robust against stroke connections/abbreviations as well as shape distortions, while maintaining reasonable robustness against stroke-order variations. Preliminary experiments are reported on tests against the Kuchibue_d-96-02 database from the Tokyo University of Agriculture and Technology. © 2001 Springer-Verlag Berlin Heidelberg.

    DOI

  • On-line Bayesian Meta-Learning: A Preliminary Report

    Y.Nakada, T.Kurihara, K.Yosui, T.Matsumoto

    CREST Workshop on Metalearning and Neuromodulation    2001

  • Pen-input on-line signature verification with position, pressure, inclination trajectories

    T. Ohishi, Y. Komiya, H. Morita, T. Matsumoto

    Proceedings - 15th International Parallel and Distributed Processing Symposium, IPDPS 2001     1757 - 1763  2001

     View Summary

    A new algorithm is proposed for pen-input on-line signature verification incorporating pen-position, pen-pressure and pen-inclinations trajectories. Preliminary experimental result looks encouraging.

    DOI

  • Bayesian MCMC nonlinear time series prediction

    Y Nakada, T Kurihara, T Matsumoto

    2001 IEEE INTERNATIONAL CONFERENCE ON ACOUSTICS, SPEECH, AND SIGNAL PROCESSING, VOLS I-VI, PROCEEDINGS   VI   3509 - 3512  2001

     View Summary

    An MCMC(Markov Chain Monte Carlo) algorithm is proposed for nonlinear time series prediction with Hierarchical Bayesian framework. The algorithm computes predictive mean and error bar by drawing samples from predictive distributions. The algorithm is tested against time series generated by (chaotic) Rossler system and it outperforms quadratic approximations previously proposed by the authors.

  • On-line signature verification algorithm incorporating pen position, pen pressure and pen inclination trajectories

    D Sakamoto, H Morita, T Ohishi, Y Komiya, T Matsumoto

    2001 IEEE INTERNATIONAL CONFERENCE ON ACOUSTICS, SPEECH, AND SIGNAL PROCESSING, VOLS I-VI, PROCEEDINGS   Ⅱ   993 - 996  2001

     View Summary

    A new algorithm is proposed for pen-input on-line signature verification incorporating pen-position, pen-pressure and pen-inclinations trajectories. Preliminary experimental result looks encouraging.

  • On-line signature verifier incorporating pen position, pen pressure and pen inclination trajectories

    H Morita, D Sakamoto, T Ohishi, Y Komiya, T Matsumoto

    AUDIO- AND VIDEO-BASED BIOMETRIC PERSON AUTHENTICATION, PROCEEDINGS   2091   318 - 323  2001

     View Summary

    This paper proposes a new algorithm PPI (pen-position/pen-pressure/pen-inclination) for on-line pen input signature verification. The algorithm considers writer's signature as a trajectory of pen-position, pen-pressure and pen-inclination which evolves over time, so that it is dynamic and biometric. Since the algorithm uses pen-trajectory information, it naturally needs to incorporate stroke number (number of pen-ups/pen-downs) variations as well as shape variations. The proposed scheme first generates templates from several authentic signatures of individuals. In the verification phase, the scheme computes a distance between the template and input trajectory. Care needs to be taken in computing the distance function because; (i) length of a pen input trajectory may be different from that of template even if the signature is genuine; (ii) number of strokes of a pen input trajectory may be different from that of template, i.e., the number of pen-ups/pen-downs obtained may differ from that of template even for an authentic signature. If the computed distance does not exceed a threshold value, the input signature is predicted to be genuine, otherwise it is predicted to be forgery. A preliminary experiment is performed on a database consisting of 293 genuine writings and 540 forgery writings, from 8 individuals. Average correct verification rate was 97.6 % whereas average forgery rejection rate was 98.7 %. Since no fine tuning was done, this preliminary result looks very promising.

  • Baysian On-line Learning: A Sequential Monte Carlo with Importance Resampling

    T.Kurihara, Y.Nakada, K.Yosui, T.Matsumoto

    NNSP2001     163 - 193  2001

  • Bayesian on-line learning: A sequential Monte Carlo with importance resampling

    T Kurihara, Y Nakada, K Yosui, T Matsumoto

    NEURAL NETWORKS FOR SIGNAL PROCESSING XI     163 - 172  2001

     View Summary

    A Bayesian on-line learning scheme with Sequential Monte Carlo incorporating Importance Resampling is proposed. The proposed scheme adjusts not only parameters for data fitting but also adjusts hyperparamaters on-line so that the scheme attempts to avoid overfitting in an adaptive manner. One of the advantages of the scheme is the fact that it can adapt to environmental changes, i. e., it can perform learning even when underlying input-output relationship varies over time. The scheme is tested against simple examples and is shown to be functional.

  • A Bayesian bi-gram scheme for HMM online handwriting recognition algorithm

    T Hasegawa, K Akamatsu, T Matsumoto

    SIXTH INTERNATIONAL CONFERENCE ON DOCUMENT ANALYSIS AND RECOGNITION, PROCEEDINGS     1012 - 1016  2001

     View Summary

    With the advent of advances in pen input devices including PDA, on-line character recognition is becoming increasingly important.
    This paper proposes a new Bayesian Bi-gram scheme for on-line character recognition using HMM.

    DOI

  • A Sequential Monte Carlo Bayesian On-line Learning

    T. Kurihara, Y. Nakada, K. Yosui, T. Matsumoto

    The 9th International Symposium on Intelligent Signal Processing and Communications Systems     381 - 385  2001

  • A Pen-input On-line Signature Verifier

    H. Morita, D. Sakamoto, T. Ohishi, T. Matsumoto

    The 9th International Symposium on Intelligent Signal Processing and Communications Systems     376 - 380  2001

  • Automatical Learning of Hyperparameters via Sequential Monte Carlo

    K. Yosui, T. Kurihara, T. Matsumoto

    International Symposium New Trends in Optimization and Computational Algorithm     121 - 122  2001

  • Multi-input floating gate differential amplifier and application to intelligent sensors

    T Sakai, H Nagai, T Matsumoto

    ANALOG INTEGRATED CIRCUITS AND SIGNAL PROCESSING   25 ( 3 ) 291 - 298  2000.12

     View Summary

    Multi-input floating gate differential amplifier (FGDA) is proposed which can perform any convolution operation with differential structure and feedback loop. All operations are in the voltage mode. Only one terminal is required for the negative feedback which can suppress distortions due to mismatches of active elements. Possible applications include intelligent image sensor, where fully parallel DCT operation can be performed. A prototype chip is fabricated which is functional. A preliminary test result is reported.

    DOI CiNii

  • A discrete HMM for online handwriting recognition

    H Yasuda, K Takahashi, T Matsumoto

    INTERNATIONAL JOURNAL OF PATTERN RECOGNITION AND ARTIFICIAL INTELLIGENCE   14 ( 5 ) 675 - 688  2000.08

     View Summary

    A fast HMM algorithm is proposed for online handwritten character recognition. After preprocessing the input strokes are discretized so that a discrete HMM is used. This particular discretization naturally leads to a simple procedure for assigning the initial state and state transition probabilities. In the training phase, complete marginelization with respect to state is not performed. A criterion based on normalized maximum likelihood ratio is given for deciding when to create a new model for the same character in the learning phase, in order to cope with stroke order variations and large shape variations. Experiments are done on the Kuchibue data base from Tokyo University of Agriculture and Technology. The algorithm appears to be very robust against stroke number variations and have reasonable robustness against stroke order variations and large shape variations. A drawback of the proposed algorithm is its memory requirement when the number of character classes and their associated model becomes large. Density tying is discussed in order to overcome this problem.

    DOI CiNii

  • A hierarchical Bayesian scheme for reconstruction and prediction of nonlinear dynamical systems

    MATSUMOTO T.

    Proc. of The International Symposium on Frontiers of Time Series Modeling   4   336 - 337  2000

    CiNii

  • Pecora-Carroll Chaotic Synchronization and Masking in an R-L-Diode Circuit

    M. Nishi, T. Matsumoto

    Electronics and Communications in Japan, Part III-Fundamental Electronics Science 2000   83 ( 2 ) 44 - 45  2000

    DOI

  • A Highly Linear Open-Loop Full CMOS High-Speed Sample-and-Hold Stage

    Kh. Hadidi, M. Sasaki, T. Watanabe, D. Muramatsu, T. Matsumoto

    IEICE Trans.Fundamentals   E83-A ( 2 ) 261 - 266  2000

  • 顔の向き推定を目指したGabor wavelet特徴量の解析

    山根智文, 松本 隆, Chippingdale Simon, 伊藤崇之

    電子情報通信学会総合大会講演論文集     226  2000

  • ペンの位置・筆圧・角度情報を利用した署名照合アルゴリズム

    大石 哲, 小宮義光, 松本 隆

    第13回回路とシステム(軽井沢)ワークショップ,論文集     71 - 76  2000

  • Chaotic Maskingにおける信号増幅作用についての考察

    永井立紀, 畠澤泰成, 藤井伊織, 松本 隆

    第13回回路とシステム(軽井沢)ワークショップ,論文集     13 - 18  2000

  • フローティング・ゲートを用いるアナログ積和演算回路のプロトタイプ・チップについて

    坂井丈泰, 広谷 俊太郎, 松本 隆

    第2回LSI IPデザイン・アワード 開発奨励賞    2000

  • 未来のイメージセンサ: ビジョンチップに挑む

    坂井 丈泰, 松本 隆

    エレクトロニクス/オーム社   45 ( 5 ) 46 - 48  2000

  • MCMC非線形カオス的時系列予測:予測分布とエラーバーの評価

    中田洋平, 松本 隆

    2000年情報論的学習理論ワークショップ予稿集     141 - 146  2000

  • 21-9 Automation of database acquisition in the FAVRET face recognition system

    CLIPPINGDALE Simon, ITO Takayuki, YAMANE Tomonori, MATSUMOTO Takashi

    PROCEEDINGS OF THE ITE ANNUAL CONVENTION   2000   308 - 309  2000

     View Summary

    The FAVRET face recognition system detects, tracks and recognizes faces in video sequences using a database computed from multiple labeled

    DOI CiNii

  • A Hierarchical Bayesian Scheme for Nonlinear Time Series Prediction Problems

    T.Matsumoto

    Pacific Rim Conference on Dynamical Systems    2000

  • A Hierarchical Bayesian Approach to Regularization Problems with Multiple Hyperparameters

    R. Takeuchi, S. Nakazawa, K. Koizumi, T. Matsumoto

    IEICE Trans Fundamentals   E-83A ( 8 ) 1641 - 1650  2000

  • オンライン手書き文字認識アルゴリズム

    小林 充, 真崎晋也, 宮本 修, 中川洋一, 小宮義光, 松本 隆

    情報処理学会論文誌   41 ( 9 ) 2536 - 2544  2000

  • オフセット補償機能を有するフローティングゲート付作動増幅器

    伊藤和則, 広谷俊太郎, 坂井丈泰, 松本 隆

    電子情報通信学会エレクトロニクスソサイエティ大会論文集   C-12-24   104  2000

  • On-line Signature Verification using Pen-Position, Pen-Pressure and Pen-Inclination Trajectories

    T. Ohishi, Y. Komiya, T. Matsumoto

    15th International Conference on Pattern Recognition     547 - 550  2000

  • Chaotic Maskingでの情報信号の増幅作用とその周波数依存性

    畠澤泰成, 永井立紀, 藤井伊織, 松本 隆

    情報処理学会全国大会論文集    2000

  • HMMとBayesian Bi-gram Schemeによるオンライン文字認識

    長谷川 智希, 松本 隆

    情報処理学会第61回全国大会論文集   2   259 - 260  2000

    CiNii

  • Bayes的非線形ダイナミカルシステムの予測:MCMCによる予測分布とエラーバーの評価

    中田 洋平, 松本 隆

    情報処理学会第61回全国大会論文集   1   197 - 198  2000

  • ペン位置・筆圧・傾き情報を用いたオンライン署名照合アルゴリズムPPI

    大石哲, 小宮義光, 松本 隆

    情報処理学会第61回全国大会論文集   4   153 - 154  2000

  • 並列抵抗回路網によるFlash-ADC性能向上の検討

    大上健志, 松本 隆

    電子情報通信学会エレクトロニクスソサイエティ大会論文集   C-12-27   107  2000

  • ビジョンチップ

    小林春夫, 松本 隆

    「映像メデイアハンドブック」の2.4.2 の項/(株)オーム社     60 - 69  2000

  • 署名照合アルゴリズムPPI

    坂本大輔, 小宮義光, 松本 隆

    電子情報通信学会情報・システムソサイエティ大会論文集   D-12-18   205  2000

  • A 430MHz, -47.5dB THD, Signal Transconductor, 5th-Order Low-Pass Filter in a 0.5μm CMOS Process

    Kh. Hadidi, 江口 啓太, 伊東 充吉, 松本 隆

    電子情報通信学会エレクトロニクスソサイエティ大会論文集   C-12-23   103  2000

  • An Analog Based Digital Adder for High Speed Applications

    Kh. Hadidi, 薊 純一郎, 松本 隆

    電子情報通信学会エレクトロニクスソサイエティ大会論文集   C-12-14   94  2000

  • A Highly Linear Open-Loop Full CMOS High-Speed Sample-and-Hold Stage

    Kh. Hadidi, 佐々木 昌浩, 渡辺 忠敏, 村松 大吾, 松本 隆

    電子情報通信学会エレクトロニクスソサイエティ大会論文集   C-12-25   105  2000

  • Quadratic Approximation vs. MCMC Predictive Mean Computations and Model Comparisons for Hierarchical Bayes Approach with Neural Nets

    Y. Nakajima, M. Asano, Y. Nakada, Y. Satoh, T. Matsumoto

    IEEE International Symposium on Intelligent Signal Processing and Communication Systems   2 ( A8-4-3 ) 765 - 768  2000

  • Nonlinear Time Series Predictions via Hierarchical Baysian MCMC: Predictive Mean and Error Bar

    Y. Nakada, T. Matsumoto

    7th International Conference on Neural Information Processing, (ICONIP)   1   199 - 204  2000

  • An On-Line Pen Input Signature Verification Algorithm

    T. Ohishi, Y. Komiya, Takashi Matsumoto

    2000 IEEE International Symposium on Intelligent Signal Processing and Communication Systems   2 ( E7-1-1 ) 589 - 592  2000

  • A Pen Input On-line Signature verifier PPI

    T. Ohishi, Y. Komiya, T. Matsumoto

    7th International Conference on Neural Information Processing, (ICONIP)   1   381 - 386  2000

  • A Fast Discrete HMM Algorithm for On-line Hand Written Character Recognition

    T. Hasegawa, T. Matsumoto

    2000 IEEE International Symposium on Intelligent Singnal Processing and Communication Systems(ISPACS)   2 ( E7-1-5 ) 609 - 612  2000

  • A Bayesian Bi-gram scheme for On-line Handwritiing Recognition with HMM

    T. Hasegawa, T. Matsumoto

    International Conference on Neural Information Processing (ICONIP)   1   392 - 397  2000

  • From Data to Nonlinear Dynamical Systems: A Hierarchical Bayesian Algorithm

    T.Matsumoto

    2nd ISM International Symposium on Frontiers of Time Series Modeling Nonparametric Approach to Knowledge Discovery    2000

  • Detecting switch dynamics in chaotic time-waveform using a parametrized family of nonlinear predictors

    Isao Tokuda, Ryuji Tokunaga, Takashi Matsumoto

    Physica D: Nonlinear Phenomena   135 ( 1-2 ) 63 - 78  2000.01

     View Summary

    An algorithm is presented for detecting switch dynamics in chaotic time-waveform. By the "switch dynamics," we mean that the chaotic time-waveform is measured from a dynamical system whose bifurcation parameters are occasionally switched among a set of slightly different parameter values. First, the switched chaotic time-waveform is divided into windows of short-term time-waveforms. From the set of windowed time-waveforms, "qualitatively similar" parametrized family of nonlinear predictors is constructed. "Qualitatively similar" parametrized family means that the family of nonlinear predictors exhibits "qualitatively similar" bifurcation phenomena as the original. By characterizing the windows of short-term chaotic time-waveforms in terms of the "qualitative" parameters of nonlinear predictors, switch dynamics of their associated bifurcation parameters are detected. For the Lorenz equations, the Rössler equations, and the Mackey-Glass equations, efficiency of the algorithm is demonstrated. In the experiment, chaotic time-waveforms contaminated with observational noise is considered. © 2000 Elsevier Science B.V.

    DOI

  • A Hierarchical Bayesian Scheme for Reconstruction and Prediction of Nonlinear Dynamical Systems

    T.Matsumoto, M. Saito, J. Sugi, Y. Nakajima

    統計数理研究所研究教育活動報告No.4   4   336 - 337  2000

  • A Hierarchical Bayesian Scheme for Nonlinear Time Series Prediction Problems

    T.Matsumoto

    Pacific Rim Conference on Dynamical Systems    2000

  • A Hierarchical Bayesian Approach to Regularization Problems with Multiple Hyperparameters

    R. Takeuchi, S. Nakazawa, K. Koizumi, T. Matsumoto

    IEICE Trans Fundamentals   E-83A ( 8 ) 1641 - 1650  2000

  • On-line Signature Verification using Pen-Position, Pen-Pressure and Pen-Inclination Trajectories

    T. Ohishi, Y. Komiya, T. Matsumoto

    15th International Conference on Pattern Recognition     547 - 550  2000

  • Fast discrete HMM algorithm for on-line handwriting recognition

    T Hasegawa, H Yasuda, T Matsumoto

    15TH INTERNATIONAL CONFERENCE ON PATTERN RECOGNITION, VOL 4, PROCEEDINGS     535 - 538  2000

     View Summary

    A fast Discrete HMM algorithm is proposed for on-line hand written character recognition. Alter preprocessing input stroke are discretized so that a discrete HMM is used. This particular discretization naturally leads to a simple procedure for assigning initial stare and state transition probabilities. In the training phase, complete marginelization with respect to state is not performed.
    A criterion based on normalized maximum likelihood ratio is given for deciding when to create a new model for the same character in the learning phase, in order to cope with stroke order variations and large shape variations.
    Experiments are done on the Kuchibue database from TUAT. The algorithm appears to be very robust against stroke number variations and have reasonable robustness against stroke order variations and large shape variations.
    A drawback of the proposed algorithm is its memory requirement when the number of character classes and their associated models becomes large.

  • Quadratic Approximation vs. MCMC Predictive Mean Computations and Model Comparisons for Hierarchical Bayes Approach with Neural Nets

    Y. Nakajima, M. Asano, Y. Nakada, Y. Satoh, T. Matsumoto

    IEEE International Symposium on Intelligent Signal Processing and Communication Systems   2 ( A8-4-3 ) 765 - 768  2000

  • Nonlinear Time Series Predictions via Hierarchical Baysian MCMC: Predictive Mean and Error Bar

    Y. Nakada, T. Matsumoto

    7th International Conference on Neural Information Processing, (ICONIP)   1   199 - 204  2000

  • An On-Line Pen Input Signature Verification Algorithm

    T. Ohishi, Y. Komiya, Takashi Matsumoto

    2000 IEEE International Symposium on Intelligent Signal Processing and Communication Systems   2 ( E7-1-1 ) 589 - 592  2000

  • A Pen Input On-line Signature verifier PPI

    T. Ohishi, Y. Komiya, T. Matsumoto

    7th International Conference on Neural Information Processing, (ICONIP)   1   381 - 386  2000

  • A Fast Discrete HMM Algorithm for On-line Hand Written Character Recognition

    T. Hasegawa, T. Matsumoto

    2000 IEEE International Symposium on Intelligent Singnal Processing and Communication Systems(ISPACS)   2 ( E7-1-5 ) 609 - 612  2000

  • A Bayesian Bi-gram scheme for On-line Handwritiing Recognition with HMM

    T. Hasegawa, T. Matsumoto

    International Conference on Neural Information Processing (ICONIP)   1   392 - 397  2000

  • Bayesian MCMC nonlinear time series prediction: Predictive mean and error bar

    Y Nakada, T Matsumoto

    NEURAL NETWORKS FOR SIGNAL PROCESSING X, VOLS 1 AND 2, PROCEEDINGS     155 - 164  2000

     View Summary

    When nonlinear dynamical system is behind time series data, predictions are rather difficult. Hierarchical Bayesian scheme previously proposed by the authors has been shown to be reasonably sound. A great difficulty implementing the Hierarchical Bayesian scheme lies in the computation of predictive distributions for which quadratic approximations have been used so far. This paper attempts to compute predictive mean and error bar for nonlinear time series prediction problems via MCMC(Markov Chain Monte Carlo) without quadratic approximations. The scheme is tested against time series generated by (Chaotic) Rossler system.

  • From Data to Nonlinear Dynamical Systems: A Hierarchical Bayesian Algorithm

    T.Matsumoto

    2nd ISM International Symposium on Frontiers of Time Series Modeling Nonparametric Approach to Knowledge Discovery    2000

  • A hierarchical Bayesian nonlinear time series prediction weighted by marginal likelihoods

    M Saito, M Asano, T Matsumoto

    NEURAL NETWORKS FOR SIGNAL PROCESSING X, VOLS 1 AND 2, PROCEEDINGS     115 - 124  2000

     View Summary

    A nonlinear time series prediction scheme is proposed with a combination of model dynamical systems weighted by model marginal likelihoods. The scheme outperforms prediction with a single model prediction with the highest marginal likelihood.

  • Multi-input floating gate differential amplifier and applications to intelligent sensors

    T Sakai, H Nagai, T Matsumoto

    IEICE TRANSACTIONS ON FUNDAMENTALS OF ELECTRONICS COMMUNICATIONS AND COMPUTER SCIENCES   E82A ( 2 ) 335 - 340  1999.02

     View Summary

    Multi-input floating gate differential amplifier (FGDA) is proposed which can perform any convolution operation with differential structure and feedback loop. All operations are in the voltage mode. Only one terminal is required for the negative feedback which can suppress distortions due to mismatches of active elements. Possible applications include intelligent image sensor, where fully parallel DCT operation can be performed. A prototype chip is fabricated which is functional. A preliminary test result is reported.

  • R-L-Diode 回路におけるTopological Horseshoeについて

    竹内規晃, 永井立紀, 松本 隆

    電子情報通信学会誌A   J 82-A ( 1 ) 21 - 30  1999

  • 非線形ダイナミカルシステムの部分システム単調性とカオス同期

    西 正信, 浜野英知, 松本 隆

    電子情報通信学会誌A   J 82-A ( 1 ) 40 - 46  1999

  • CMOS 並列画像フィルタプロトタイプチップの評価

    国島貴志, 永井宏昌, 坂井丈泰, 松本 隆

    電気学会研究会資料, 電子回路研究会   ECT-99-15   13 - 18  1999

  • A Discrete HMM for On-Line Hand Writing Recognition

    H. Yasuda, K. Takahashi, T. Matsumoto

    The Second International Conference on Multimodal Interface   ( III ) 6 - 11  1999

  • Chaotic Masking における情報信号の増幅作用

    永井立紀, 浜野英知, 吉田倫己, 相田能之, 松本隆

    電気学会情報処理研究会資料   IP-99-1   1 - 6  1999

  • Chaotic Maskingにおける情報信号の増幅作用 I :R-L-Diode回路

    浜野英知, 松本 隆

    電子情報通信学会総合大会講演論文集   A-2-37   101  1999

  • Chaotic Maskingにおける情報信号の増幅作用III : 一般化Hopfield回路

    相田能之, 吉田倫己, 松本 隆

    電子情報通信学会総合大会講演論文集   A-2-39   103  1999

  • 非線形時系列のモデル周辺尤度による重み付き予測

    斉藤幹貴, 松本 隆

    電子情報通信学会技術研究報告   NC98-115   129 - 136  1999

  • カオス時系列のモデル周辺尤度による重み付き予測

    斉藤幹貴, 松本 隆

    電子情報通信学会総合大会講演論文集   D-2-33   40  1999

  • 非線形ダイナミカルシステムと再構成と予測

    松本 隆

    平成10年度総合研究大学院大学国際シンポジウム・複雑系への戦略、構成と記述     36 - 40  1999

  • ペン位置・筆圧・傾き情報を用いたオンライン署名照合

    小宮義光, 松本 隆

    電子情報通信学会総合大会講演論文集   D-12-44   217  1999

  • 顔画像からの顔のポーズ推定方法

    山根智文, 松本 隆, サイモン・クリビングデル, 伊藤崇之

    電子情報通信学会総合大会講演論文集   D-12-88   261  1999

  • Deconvolution問題の階層Bayes的アプローチ:GaN Photoliuminescence

    佐藤隆元, 岡田栄仁, 松井淳, 松本隆

    電子情報通信学会信学技報   NC 98-121   177 - 184  1999

  • DCT 演算画像フィルタプロトタイプチップの評価

    国島貴志, 永井宏昌, 坂井丈泰, 広谷俊太郎, 松本 隆

    電子情報通信学会総合大会講演論文集   C-12-57   155  1999

  • Lai-Grebogi同期とChaotic Masking

    吉田倫己, 松本 隆

    電子情報通信学会総合大会講演論文集   A-2-38   102  1999

  • Chaotic Maskingにおける情報信号の増幅作用II:Lorenz系

    永井立紀, 松本 隆

    電子情報通信学会総合大会講演論文集   A-2-36   100  1999

  • Reconstruction and Prediction of Nonlinear Dynamical Systems :A Hierarchical Bayes Approach with Nueral Nets

    T. Matsumoto, M. Saito, Y. Nakajima, J. Sugi, H. Hamagishi

    IEEE International Confence on Accoustics, Speech, and Signal Processing (ICASSP 99)   2   1057 - 1060  1999

  • Spatial and Temporal Dynamics of Vision Chips Including Parasitic Inductances and Capacitances

    H. Kobayashi, T. Matsumoto

    IEICE, Trans. Fundamentals   E82-A ( 3 ) 412 - 416  1999

  • Chaotic Masking における情報信号の増幅作用

    永井立紀, 浜野英知, 吉田倫己, 相田能之, 松本 隆

    第12回回路とシステムワークショップ論文集     7 - 12  1999

  • 103 MHz THD-61dB Full CMOS 開ループサンプル/ホールド回路

    Khayrolla Hadidi, 佐々木昌浩, 渡辺忠敏, 村松大吾, 薊純一郎, 松本 隆

    第12回回路とシステムワークショップ論文集     43 - 47  1999

  • 2次同相写像におけるトラッピング領域の存在証明

    蟹江幸司, 徳永隆治, 松本 隆

    電子情報通信学会誌   J82-A ( 5 ) 619 - 626  1999

  • データから非線形ダイナミクスへ -階層ベイズ的時系列予測-

    松本 隆, 浜岸 広明, 杉 淳二郎, 斉藤 幹貴, 長南 吉正

    電子情報通信学会論文誌DII   J82-D-II ( 6 ) 1059 - 1071  1999

  • 階層ベイズ的ニューラルネットのモデル比較および予測分布:二次近似 vs. MCMC

    中島芳徳, 浅野正登, 中田洋平, 松本 隆

    電子情報通信学会技術研究報告   NC 99-35, vol. 99 ( 193 ) 49 - 54  1999

  • 異なる電源によるR-L-Diode回路のPecora-Carrollカオス同期とマスキング

    西 正信, 浜野英知, 松本 隆

    電子情報通信学会論文誌   J82-A ( 7 ) 1160 - 1161  1999

  • Nonlinear Time Series Prediction Weighted by Marginal Likelihoods: A Hierarachical Bayestian Approach

    T. Matsumoto, M. Saito, J. Sugi

    IJCNN '99   4 ( 572 ) 2604 - 2607  1999

  • Selection of Model Dynamical Systems for Nonlinear Time Series Prediction Problems with Hierarchical Bayesisan Neural Nets

    T. Matsumoto, J. Sugi, M. Saito, Y. Nakajima, H. Hamagishi

    1999年情報論的学習理論ワークショップ予稿集(IBIS'99)     201 - 206  1999

  • Quadratic Approximation vs. MCMC for Hierarchical Bayesian NeuralNets: Model Comparisons and Predictive Mean Computatiaons

    M. Asano, Y. Nakada, Y. Nakajima, T. Matsumoto

    1999年情報論的学習理論ワークショップ予稿集(Proceedings of 1999 Workshop on Information-Based Induction Sciences (IBIS'99)   pp.117-121   117 - 121  1999

  • 階層ベイズ的ニューラルネットの二次近似法・MCMC法によるモデル比較と予測分布

    中島芳徳, 浅野正登, 中田洋平, 松本 隆

    日本神経回路学会第9回全国大会論文集   P3-16   185 - 186  1999

  • 最小埋め込み次元の推定ー階層ベイズ的手法とFNN法

    杉 淳二郎, 松本 隆

    日本神経回路学会第9回全国大会論文集   P3-15   183 - 184  1999

  • Deconvolution 問題の階層Bayes的アプローチ:GaN photoluminescence

    佐藤隆元, 広畠 勉, 松本 隆

    日本神経回路学会第9回全国大会論文集   P3-17   187 - 188  1999

  • ニューロンMOSトランジスタによるアナログ積和演算回路チップの評価

    広谷俊太郎, 国島貴志, 永井宏昌, 坂井丈泰, 松本 隆

    日本神経回路学会第9回全国大会論文集   P1-27   67 - 68  1999

  • Nonlinear Dynamical Systems Approach to Building Energy Prediction Problems

    NAKAJIMA Y.

    Proc. Building Simulation'99, Kyoto, Japan   II   901 - 907  1999

    CiNii

  • A 430MHz, -52 dB THD, Single Transconductor, 3rd-Order Low-Pass Filters and its Extension to a 5th-Order, in a 0.5 μm CMOS Process

    Kh. Hadidi, K. Eguchi, T. Matsumoto

    25th European Solid-State Circuits Conference (ESSCIRC '99)     390 - 393  1999

  • A 500MS/sec-54dB THD S/H Circuit in a 0.5 μm CMOS Process

    Kh. Hadidi, D. Muramatsu, T. Oue, T. Matsumoto

    25th European Solid-State Circuits Conference (ESSCIRC '99)     158 - 161  1999

  • Discrete HMM ペン入力オンライン文字認識

    長谷川智希, 安田英史, 高橋健一郎, 松本 隆

    ヒューマンインタフェイスシンポジウム'99論文集   1113   11 - 16  1999

  • ペン位置、筆圧、傾き情報を用いたオンライン署名照合アルゴリズムPPI

    大石 哲, 小宮義光, 松本 隆

    ヒューマンインタフェイスシンポジウム'99論文集   1112   5 - 10  1999

  • エッジ検出用インテリジェントイメージセンサチップの試作

    広谷俊太郎, 坂井丈泰, 渡部俊久, 松本 隆

    電子情報通信学会技術研究報告、EID99-53     39 - 44  1999

  • On-line Pen Input Signature Verification PPI (pen-Position/pen-Pressure/pen-Inclination)

    Y. Komiya, T. Matsumoto

    IEEE SMC'99 (IEEE International Conference onSystems, Man, and Cybernetics)   IV   41 - 46  1999

  • A Hierarchical Bayesian Scheme for Nonlinear Dynamical System Reconstruction and Prediction with Neural Nets

    MATSUMOTO T.

    Proc. 1999 IEEE International Conference on Systems, Man, and Cybernetics, Tokyo, Japan   IV   1119 - 1124  1999

    CiNii

  • 階層ベイズ的空調機熱負荷予測--ニューラルネットによる非線形ダイナミカルシステム的アプローチ--

    中島芳徳, 斉藤幹貴, 杉淳二郎, 浜岸広明, 松本 隆

    電子情報通信学会論文誌D-II   J82-D-II ( 11 ) 2075 - 2083  1999

  • On-line pen input signature verifier PPI (pen-position/pen-pressure/pen-inclination)

    Y. Komiya, T. Matsumoto

    ICONIP 1999, 6th International Conference on Neural Information Processing - Proceedings   3   1236 - 1240  1999

     View Summary

    An algorithm is proposed for pen-input online signature verification, incorporating pen-position, pen-pressure and pen-inclination trajectories. The algorithm considers the writer's signature as a trajectory of pen-position, pen-pressure and pen-inclination which evolves over time, so that it is dynamic and biometric. Since the algorithm uses pen-Trajectory information, it naturally needs to incorporate stroke number (number of pen-ups/pen-downs) variations as well as shape variations. The proposed scheme first generates templates from several authentic signatures of individuals. In the verification phase, the scheme computes a distance between the template and an input trajectory. Care needs to be taken in computing the distance function because: (i) the length of a pen input trajectory may be different from that of a template even if the signature is genuine
    (ii) the number of strokes of a pen input trajectory may be different from that of the template, i.e., the number of pen-ups/pen-downs obtained may differ from that of the template even for an authentic signature. If the computed distance does not exceed a threshold value, the input signature is predicted to be genuine, otherwise it is predicted to be forgery. Preliminary experimental results look encouraging.

    DOI

  • ビジョンチップ

    小林春夫, 松本 隆

    第7回光インターコネクト情報処理研究会, 第89回光コンピューティング研究会資料   OIP 99-17   25 - 31  1999

  • エッジ検出用ビジョンチップの試作

    伊藤和則, 広谷俊太郎, 坂井丈泰, 渡部俊久, 松本 隆

    映像情報メディア学会冬季大会講演予稿集   3 ( 3 ) 60  1999

  • A Discrete HMM for On-Line Hand Writing Recognition

    H. Yasuda, K. Takahashi, T. Matsumoto

    The Second International Conference on Multimodal Interface   ( III ) 6 - 11  1999

  • Subsystem decreasing for exponential synchronization of chaotic systems

    Takashi Matsumoto, Masanobu Nishi

    Physical Review E - Statistical Physics, Plasmas, Fluids, and Related Interdisciplinary Topics   59 ( 2 ) 1711 - 1718  1999

     View Summary

    Conditions are derived under which a general class of nonlinear dynamical systems admits chaotic synchronization. The result is applied to the Lorenz system, Rössler’s second equation, a generalized Hopfield network, and a driven [Formula Presented]-diode circuit. Several experimental as well as numerical results are also given to confirm the theory. © 1999 The American Physical Society.

    DOI CiNii

  • Reconstruction and Prediction of Nonlinear Dynamical Systems :A Hierarchical Bayes Approach with Nueral Nets

    T. Matsumoto, M. Saito, Y. Nakajima, J. Sugi, H. Hamagishi

    IEEE International Confence on Accoustics, Speech, and Signal Processing (ICASSP 99)   2   1057 - 1060  1999

  • Spatial and Temporal Dynamics of Vision Chips Including Parasitic Inductances and Capacitances

    H. Kobayashi, T. Matsumoto

    IEICE, Trans. Fundamentals   E82-A ( 3 ) 412 - 416  1999

  • A hierarchical Bayes approach to reconstruction and prediction of nonlinear dynamical systems

    T Matsumoto, M Saito, Y Nakajima, J Sugi, H Hamagishi

    PROCEEDINGS OF THE IEEE-EURASIP WORKSHOP ON NONLINEAR SIGNAL AND IMAGE PROCESSING (NSIP'99)   1   114 - 118  1999

     View Summary

    An attempt is made. to solve two classes of nonlinear time series prediction problems with a hierarchical Bayes Approach using neural nets.

  • Nonlinear Time Series Prediction Weighted by Marginal Likelihoods: A Hierarachical Bayestian Approach

    T. Matsumoto, M. Saito, J. Sugi

    IJCNN '99   4 ( 572 ) 2604 - 2607  1999

  • Selection of Model Dynamical Systems for Nonlinear Time Series Prediction Problems with Hierarchical Bayesisan Neural Nets

    T. Matsumoto, J. Sugi, M. Saito, Y. Nakajima, H. Hamagishi

    1999年情報論的学習理論ワークショップ予稿集(IBIS'99)     201 - 206  1999

  • Quadratic Approximation vs. MCMC for Hierarchical Bayesian NeuralNets: Model Comparisons and Predictive Mean Computatiaons

    M. Asano, Y. Nakada, Y. Nakajima, T. Matsumoto

    1999年情報論的学習理論ワークショップ予稿集(Proceedings of 1999 Workshop on Information-Based Induction Sciences (IBIS'99)   pp.117-121   117 - 121  1999

  • Nonlinear Dynamical Systems approach to Building Energy Prediction Problems

    Y. Nakajima, M. Saito, J. Sugi, T. Matsumoto

    Proceedings of Building Simulation '99, Sixth International IBPSA Conference   II   901 - 907  1999

  • A 430MHz, -52 dB THD, Single Transconductor, 3rd-Order Low-Pass Filters and its Extension to a 5th-Order, in a 0.5 μm CMOS Process

    Kh. Hadidi, K. Eguchi, T. Matsumoto

    25th European Solid-State Circuits Conference (ESSCIRC '99)     390 - 393  1999

  • A 500MS/sec-54dB THD S/H Circuit in a 0.5 μm CMOS Process

    Kh. Hadidi, D. Muramatsu, T. Oue, T. Matsumoto

    25th European Solid-State Circuits Conference (ESSCIRC '99)     158 - 161  1999

  • On-line Pen Input Signature Verification PPI (pen-Position/pen-Pressure/pen-Inclination)

    Y. Komiya, T. Matsumoto

    IEEE SMC'99 (IEEE International Conference onSystems, Man, and Cybernetics)   IV   41 - 46  1999

  • A Hierarchical Bayesian Scheme for Nonlinear Dynamical SystemReconstruction and Prediction with Neural Nets

    T. Matsumoto, Y. Nakajima, M. Saito, J. Sugi

    IEEE SMC'99 (IEEE International Conference on Systems, Man, and Cybernetics)   IV   1119 - 1124  1999

  • Model comparisons and predictive mean computations for hierarchical Bayesian neural nets: Quadratic approximation vs. MCMC

    Y. Nakajima, M. Asano, Y. Nakada, T. Matsumoto

    ICONIP 1999, 6th International Conference on Neural Information Processing - Proceedings   1   137 - 141  1999

     View Summary

    The article is a first step toward an attempt to demonstrate the validity of quadratic approximations (QAP) of computing marginal likelihood as well as predictive distributions for the hierarchical Bayesian scheme by using MCMC (Markov chains Monte Carlo). At least for the simple examples considered, the QAP gives reasonable results for marginal likelihood and predictive distributions. More elucidation is necessary to further study the issues for more complicated problems including nonlinear time series prediction problems.

    DOI

  • A hierarchical Bayesian deconvolution with positivity constraints

    T. Satoh, A. Matsui, T. Hirohata, T. Matsumoto

    ICONIP 1999, 6th International Conference on Neural Information Processing - Proceedings   3   1230 - 1235  1999

     View Summary

    A class of deconvolution problems with positivity constraints is formulated in terms of a hierarchical Bayesian framework. A deconvolution algorithm is proposed and applied to a specific real world problem: estimation of relaxation dynamics of GaN photoluminescence (S. Nakamura et al., 1994).

    DOI

  • On-line Pen Input Signature Verifer PPI (pen-Position/pen-Pressure/pen-Inclination)

    M. Komiya, T. Matsumoto

    ICONIP '99 (6th International Conference on Nueral Information Processing)   III   1236 - 1240  1999

  • A floating-gate MOS implementation of resistive fuse

    T Sawaji, T Sakai, H Nagai, T Matsumoto

    NEURAL COMPUTATION   10 ( 2 ) 485 - 498  1998.02

     View Summary

    Resistive fuses are key elements in weak string filters, which smooth out noise while detecting step edges inherent in original data. A resistive fuse is implemented by two pairs of floating-gate metal oxide semiconductor (MOS) transistors in a chip by a standard double poly complementary MOS process.

    DOI CiNii

  • オンライン手書き文字認識アルゴリズムRAV

    真崎晋也, 松本 隆

    ペン入力研究談話会    1998

  • Neural Netによる非線形ダイナミカルシステムの再構成と予測

    松本隆

    統計数理研究所研究会, 重回帰分析・ニューラルネットワークにおける変数選択とベイズ的方法    1998

  • HMMを用いたオンライン手書き文字認識

    安田英史, 松本 隆

    ペン入力研究談話会    1998

  • フローティングMOSを用いたResistive Fuseの実装

    澤地利明, 坂井丈泰, 永井宏昌, 松本 隆

    映像情報メディア学会誌   52 ( 2 ) 200 - 205  1998

    DOI CiNii

  • データからダイナミクスへ:階層ベイズ的非線形時系列予測

    浜岸広明, 杉淳一郎, 斉藤幹貴, 松本 隆

    電子情報通信学会技術研究報告   97, pp. 73 - 80, ( 530, NLP-97-133 ) 73 - 80  1998

  • Air-conditioning Load Prediction by Hierarchical Bayesian Neural Nets

    NAKAJIMA Y., SAITO M., SUGI J., HAMAGISHI H., HATTORI D., MATSUMOTO T.

    IEICE technical report. Nonlinear problems   97 ( 530,NLP-135 ) 89 - 96  1998

     View Summary

    Hierarchical Bayesian neural nets are formulated to energy demand prediction, and are applied to air-conditioning load prediction.

    CiNii

  • 非線形ダイナミカルシステムの部分システム単調性とカオス同期

    西 正信, 松本 隆

    電子情報通信学会技術研究報告   97 ( 530,NLP-139 ) 25 - 32  1998

  • Hierarchical Bayesian Deconvolution : Estimation of GaN Radiation/Relaxation Dynamics

    SATOH T., OKADA E., MATSUI A., MATSUMOTO T.

    IEICE technical report. Nonlinear problems   97 ( 530,NLP-134 ) 81 - 88  1998

     View Summary

    When a strong laser is projected onto a semiconductor material, photoluminescence is observed. This process is important to understand opto-electronic properties of semiconductor material. A class of deconvolution problems with positivity constraints is formulated in terms of a hierarchical Bayes framework. The proposed algorithm is applied to estimation of photoluminescence dynamics structure which is becoming extremely popular for its blue laser emission.

    CiNii

  • 表面反射特性及び3次元形状の知覚に及ぼす陰影の効果

    樋口喜昭, 藤井真人, 伊藤崇之, 放送技研, 松本 隆

    電子情報通信学会技術研究報告   97 ( 598 ) 25 - 32  1998

  • Vision Chip Architecture with Light Adaptation Mechanism

    T. Yagi, H. Kobayashi, T. Matsumoto, K. Tanaka

    Artificial Life Robotics (1998)   2 ( 1 ) 12 - 18  1998

    DOI CiNii

  • Parallel Analog Image Processing : Solving Regularization Problems with Architecture Inspired by the Vertebrate Retinal Circuit

    T. Yagi, H. Kobayashi, T. Matsumoto

    Image Processing and Pattern Recognition, Academic Press     201 - 285  1998

  • 600MHz高線形3次および5次ローパスフィルタ松本 隆、江口啓太

    松本 隆, 江口啓太

    日経BP社 IP アワード賞    1998

  • Hierarchical Bayesian Neural Nets for Air-conditioning Load Prediction: Nonlinear Dynamics Approach

    Y. Nakajima, J. Sugi, M. Saito, H. Hamagishi, D. Hattori, T.Matsumoto

    International Joint Conference on Neural Networks(IJCNN '98)     1948 - 1953  1998

  • From Data to Dynamics : Predicting Chaotic Time Series by Hierarchical Bayesian Neural Nets

    T. Matsumoto, H. Hamagishi, J. Sugi, M. Saito

    International Joint Conference on Neural Networks (IJCNN '98),     2535 - 2538  1998

  • Spatial and temporal stability of vision chips including parasitic inductances and capacitances

    Haruo Kobayashi, Takashi Matsumoto

    ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings   2   1081 - 1084  1998

     View Summary

    There are two dynamics issues in vision chips: (i) the temporal dynamics issue due to the parasitic capacitors in a CMOS chip, and (ii) the spatial dynamics issue due to the regular array of processing elements in a chip. In this paper we consider parasitic inductances as well as parasitic capacitances for a network dynamics model. We show that in some cases the temporal stability condition for the network with parasitic inductances and capacitances is equivalent to that for the network with only parasitic capacitances, but in general they are not equivalent. We also show that the spatial stability conditions are equivalent in both cases. © 1998 IEEE.

    DOI

  • DCTも可能なアナログ演算回路のためのプロトタイプチップの評価

    国島貴志, 永井宏昌, 坂井丈泰, 松本 隆

    映像情報メディア学会年次大会講演予稿集   10 ( 1 ) 138 - 139  1998

  • 階層Bayes的熱負荷予測:データから非線形ダイナミカルシステムへ

    中島芳徳, 斉藤幹貴, 杉淳二郎, 松本 隆

    空気調和・衛生工学会, 学術講演会講演論文集   1   181 - 184  1998

  • From Data to Nonlinear Dynamics : A Hierarchical Bayes Approach to Neural Networks

    T. Matsumoto, Y. Nakajima, H.Hamagishi, J. Sugi, M. Saito

    IEEE Workshop on Neural Nets for Signal Processisng     333 - 342  1998

  • On-line Handwriting Recognition by Discrete HMM with Fast Learning

    H. Yasuda, K. Takahashi, T. Matsumoto

    The 6th International Workshop on Frontiers in Handwriting Recognition     15 - 24  1998

  • 非線形ダイナミカルシステムの再構成と予測:ニューラルネット的アプローチ

    松本 隆

    日本応用数理学会1998年度年会講演予稿集     184 - 185  1998

  • 非線形ダイナミカルシステムの再構成と予測

    松本 隆, 中島芳徳, 杉淳二郎, 斎藤幹貴

    平成10年度統計数理研究所「逆問題とその周辺(5)」共同研究レポート122     3 - 10  1998

  • A Novel Design Technique for Imput Differential Pairs in Single-Ended Operational Amplifiers

    M. Morimoto, Kh. Hadidi, K. Futami, T. Matsumoto

    International Conference on Electronics, Circuits and Systems(ICECS)   3   365 - 368  1998

  • A highly linear second-order stage for 500-MHz third-order and fifth-order filters

    HADIDI KH.

    IEEE International Conference on Electronics, Circuits, and Systems, Sept. 1998   3   361 - 364  1998

    CiNii

  • A Novel Highly Linear CMOS Buffer

    Kh. Hadidi, J. Sobhi, A. Hasankhaan, D. Muramatsu, T. Matsumoto

    International Conference on Electronics, Circuits and Systems(ICECS)   3   369 - 371  1998

  • A Hierarchical Bayes Algorithm for Air-Conditioning Load Prediction:Nonlinear Dynamics Approach

    Y. Nakajima, J. Sugi, M. Saito, H. Hamagishi, T. Matsumoto

    ICONIP/JNNS'98     1347 - 1350  1998

  • Chaotic Time Series Prediction via Hierarchical Bayesian Nueral Net

    T. Matsumoto, H. Hamagishi, J. Sugi, M. Saito

    ICONIP/JNNS'98     1020 - 1023  1998

  • Vision Chip Architecture with Light Adaptation Mechanism

    T. Yagi, H. Kobayashi, T. Matsumoto, K. Tanaka

    Artificial Life Robotics (1998)   2 ( 1 ) 12 - 18  1998

    DOI CiNii

  • Parallel Analog Image Processing : Solving Regularization Problems with Architecture Inspired by the Vertebrate Retinal Circuit

    T. Yagi, H. Kobayashi, T. Matsumoto

    Image Processing and Pattern Recognition, Academic Press     201 - 285  1998

  • Hierarchical Bayesian Neural Nets for Air-conditioning Load Prediction: Nonlinear Dynamics Approach

    Y. Nakajima, J. Sugi, M. Saito, H. Hamagishi, D. Hattori, T.Matsumoto

    International Joint Conference on Neural Networks(IJCNN '98)     1948 - 1953  1998

  • From Data to Dynamics : Predicting Chaotic Time Series by Hierarchical Bayesian Neural Nets

    T. Matsumoto, H. Hamagishi, J. Sugi, M. Saito

    International Joint Conference on Neural Networks (IJCNN '98),     2535 - 2538  1998

  • An open-loop full CMOS 103MHz -61dB THD S/H circuit

    K Hadidi, M Sasaki, T Watanabe, D Muramatsu, T Matsumoto

    IEEE 1998 CUSTOM INTEGRATED CIRCUITS CONFERENCE - PROCEEDINGS     381 - 383  1998

     View Summary

    Based on a real open loop architecture and a cascode-driver CMOS source-follower, we implemented a S/H circuit in a 0.8 mu m digital CMOS process. The circuit achieved -61dB THD at a sampling rate of 103MHz, while a 1.42V(p-p) 10MHz input signal was applied. This includes all parastic loading and transient effect.

  • Spatial and temporal stability of vision chips including parasitic inductances and capacitances

    H Kobayashi, T Matsumoto

    PROCEEDINGS OF THE 1998 IEEE INTERNATIONAL CONFERENCE ON ACOUSTICS, SPEECH AND SIGNAL PROCESSING, VOLS 1-6   2   1081 - 1084  1998

     View Summary

    There are two dynamics issues in vision chips: (i) The temporal dynamics issue due to the parasitic capacitors in a CMOS chip, and (ii) the spatial dynamics issue due to the regular array of processing elements in a chip. These issues are discussed in [1, 2, 3] for the resistor network with only associated parasitic capacitances. However, in this paper we consider also parasitic inductances as well as parasitic capacitances for a more precise network dynamics model. We show that in some cases the temporal stability condition for the network with parasitic inductances and capacitances is equivalent to that for the network with only parasitic capacitances, but in general they are not equivalent. We also show that the spatial stability conditions are equivalent in both cases.

    DOI

  • From Data to Nonlinear Dynamics : A Hierarchical Bayes Approach to Neural Networks

    T. Matsumoto, Y. Nakajima, H.Hamagishi, J. Sugi, M. Saito

    IEEE Workshop on Neural Nets for Signal Processisng     333 - 342  1998

  • On-line Handwriting Recognition by Discrete HMM with Fast Learning

    H. Yasuda, K. Takahashi, T. Matsumoto

    The 6th International Workshop on Frontiers in Handwriting Recognition     15 - 24  1998

  • A Novel Design Technique for Imput Differential Pairs in Single-Ended Operational Amplifiers

    M. Morimoto, Kh. Hadidi, K. Futami, T. Matsumoto

    International Conference on Electronics, Circuits and Systems(ICECS)   3   365 - 368  1998

  • A Highly Linear Second-Order Stage for 500-MHz Third-Order and Fifth-Order Filters

    Kh. Hadidi, K. Eguchi, T. Matsumoto, H. Kobayashi

    International Conference on Electronics, Circuits and Systems(ICECS)   3   361 - 364  1998

  • A Novel Highly Linear CMOS Buffer

    Kh. Hadidi, J. Sobhi, A. Hasankhaan, D. Muramatsu, T. Matsumoto

    International Conference on Electronics, Circuits and Systems(ICECS)   3   369 - 371  1998

  • A nonlinear prediction technique for parametrized families of chaotic dynamics

    R Tokunaga, T Matsumoto

    INTERNATIONAL JOURNAL OF INTELLIGENT SYSTEMS   12 ( 4 ) 291 - 309  1997.04

     View Summary

    A simple algorithm is proposed for reconstruction of parametrized families of chaotic dynamics. This algorithm enables one to generate bifurcation diagrams which are qualitatively the same as the original ones only from several time-waveforms, without knowing an explicit form of the dynamics and information of the parameter values. The algorithm consists of two steps. First, globally smooth nonlinear predictors are computed for all time waveforms. Second, the Karhunen-Loeve transform is used to find only significant parameters contributing to the bifurcations. The algorithm is tested against two parametrized families of dynamics: the Henon family and the coupled logistic/delayed-logistic family. (C) 1997 John Wiley & Sons, Inc.

    DOI CiNii

  • 階層ベイズ的アプローチによるDeconvolution

    佐藤隆元, 岡田栄仁, 松井 淳, 松本 隆

    1997年電子情報通信学会基礎・境界ソサイエティ大会講演論文集   A-4-21   91  1997

  • ダイナミカルノイズに隠された力学系とその分岐構造の推定

    手賀俊行, 徳永隆治, 梶原志保子, 松本 隆

    電子情報通信学会論文誌   J80-A ( 1 ) 91 - 104  1997

  • 非線形時系列の階層Bayes的アプローチ

    浜岸広明, 松本 隆

    電子情報通信学会技術研究報告   NLP96-131   55 - 62  1997

  • A Comparison of Laplacian Prior and Gaussian Prior for Hierarchical Bayes Learning in Neural Nets

    CHONAN Y., MATSUMOTO T.

    IEICE technical report. Nonlinear problems   NC-96-86 ( 509 ) 63 - 70  1997

     View Summary

    A hierarchical Bayes approach is taken in training Feedforward Neural Nets. Assuming Gaussian distribution and Laplacian distribution as prior probability of weight parameters, both models are evaluated by comparing their marginal likelihood. Simulation result shows marginal likelihood has close contacts with generalization ability.

    CiNii

  • M.L.P.におけるARD(Automatic Relevance Determination)の階層Bayes的アプローチ

    中島芳徳, 松本 隆

    電子情報通信学会技術研究報告NLP96-149   96 ( 510 ) 63 - 70  1997

  • インテリジェントイメージセンサのためのフローティングを用いる演算方式について

    坂井丈泰, 松本 隆

    映像情報メディア学会誌   51 ( 2 ) 263 - 269  1997

    DOI CiNii

  • 階層ベイズ的Deconvolution について

    岡田栄仁, 松井 淳, 松本 隆

    電子情報通信学会技術報告、NLP96-148   96 ( 510 ) 55 - 62  1997

  • R-L-Diode回路におけるPecora-Carrol同期回路の単純化

    浜野英知, 西 正信, 松本 隆

    1997年電子情報通信学会総合大会   A-2-46   103  1997

  • オンライン手書き文字認識アルゴリズムRAVの筆順違い文字自動登録

    真樹晋哉, 小林充, 宮本修, 中川洋一, 松本 隆

    電子情報通信学会、パターン認識メディア理解研究会   PRMU 96-210   135 - 142  1997

  • 階層ベイズ的標準正則化における対数尤度のHassian について

    松本素明, 中沢進, 松本 隆

    1997年電子情報通信学会総合大会   A-4-1   132  1997

  • 電子回路に見るカオス、分岐、同期

    松本 隆, 西 正信

    電子情報通信学会技術報告   ED-96-216   1 - 8  1997

  • フローティングゲートを用いるアナログ積和演算回路の演算精度向上について

    坂井丈泰, 永井宏昌, 松本 隆

    電子情報通信学会総合大会講演論文集     204  1997

  • Hidden Markov Model を用いたオンライン手書き文字認識

    高橋賢一郎, 安田英史, 松本 隆

    電子情報通信学会技術報告   PRMU 96-211   143 - 150  1997

  • CMOS Floating Gate Resistive Fuse Chip

    NAGAI Hiromasa, SAKAI Takeyasu, SAWAJI Toshiaki, MATSUMOTO Takashi

    IEICE technical report. Neurocomputing   NC 96-113 ( 583 ) 1 - 7  1997

     View Summary

    Weak String is a nonlinear parallel filter which can smooth out noise while it detects step edges inherent to original data. Nonlinear elements called resistive fuses are necessary for implementing weak string. This paper shows a chip implementation of resistive fuse by four floating gate CMOS transistors.

    CiNii

  • フローティングゲートを用いるアナログ積和演算回路とその応用について

    坂井丈泰, 松本 隆

    第10回回路とシステム軽井沢ワークショップ論文集     41 - 46  1997

  • 正値性が付帯するDeconvolutionのベイズ的アプローチ

    岡田栄仁, 松井淳, 佐藤隆元, 松本隆

    第10回回路とシステム軽井沢ワークショップ論文集     273 - 278  1997

  • フローティングゲート構造を用いたCMOS Resistive Fuse チップ

    永井宏昌, 坂井丈泰, 澤地利明, 松本 隆

    第10回回路とシステム軽井沢ワークショップ論文集     35 - 40  1997

  • A Hierarchical Bayes Approach to Nonlinear Time Series Prediction with Neural Nets

    T. Matsumoto, H. Hamagishi, Y. Chonan

    ICNN'97   4   2028 - 2033  1997

  • フローティングゲートを用いるアナログ積和演算回路のプロトタイプチップについて

    坂井丈泰, 永井宏昌, 国島貴志, 松本 隆

    電子情報通信学会技術研究報告   97 ( 230 ) 79 - 85  1997

  • R-L-Diode 回路におけるTopological Horseshoe: A Computer Assisted Proof Topological Horseshe in the R-L-Diode circuit

    竹内規晃, 松本 隆

    電気学会研究会、情報処理研究会   IP-97-13   1 - 10  1997

  • R-L-Diode回路におけるTopological Horseshoe: Computer Assisted Proof

    竹内規晃, 松本 隆

    電子情報通信学会基礎・境界ソサイエティ大会   A-2-1   29  1997

  • 標準正則化のハイパラメータおよびRegularizerの推定法について

    竹内 亮, David J. C. MacKay, 中沢 進, 松本 隆

    電子情報通信学会誌   J80-DII ( 9 ) 2502 - 2511  1997

  • R-L-Diode回路のPecora-Carroll カオス同期とマスキング

    西 正信, 松本 隆

    電子情報通信学会論文誌A   J80-A ( 9 ) 1421 - 1430  1997

  • R-L-Diode回路のPecora-Carroll カオス同期とマスキング

    西 正信, 松本 隆

    電子情報通信学会基礎・境界ソサイエティ大会   A-2-2   30  1997

  • フローティングゲートを用いるアナログ積和演算回路の集積度について

    坂井丈泰, 国島貴志, 永井宏昌, 松本 隆

    1997年電子情報通信学会ソサイエティ大会講演論文集、エレクトロニクス2     121  1997

  • Threshold Controllable COMS Resistive Fuse Chip

    NAGAI Hiromasa, SAKAI Takeyasu, SAWAJI Toshiaki, MATSUMOTO Takashi

    ITE Technical Report   21 ( 74 ) 1 - 6  1997

     View Summary

    Weak string filter which can smooth out noise while it detects and preserves step edges inherent to original data needs nonlinear elements called resistive fuses. This paper shows a chip implementation of resistive fuse by two pairs of floating gate CMOS transistors where threshold parameters are controllable.

    DOI CiNii

  • A Hierarchical Bayesian Deconvolution with Positivity Constraints

    T.Satoh, A.Matsui, E.Okada, T.Matsumoto

    Proceedings of the 29th ISCIE International Symposium Stochastic Systems Theory and Applications     195 - 200  1997

  • Nonlinear Time Series Predictions via Hierarchical Bayes approach

    T. Matsumoto, H. Hamagishi, Y. Chonan

    Proceedings of the 29th ISCIE International Symposium Stochastic Systems Theory and Applications     239 - 244  1997

  • A Hierarchical Bayesian Regularizer Comparison for Regularization Problems

    S. Nakazawa, R. Takeuchi, T. Matsumoto

    Proceedings of the 29th ISCIE International Symposium Stochastic Systems Theory and Applications     115 - 120  1997

  • 分岐図再構成 ― 時系列信号からの係数族推定

    徳永隆治, 松本 隆, 徳田功

    応用数理   7 ( 4 ) 17 - 28  1997

  • Implementing resistive fuse with floating gate MOS transistors

    T Sawaji, T Sakai, H Nagai, T Kunishima, T Matsumoto

    1997 IEEE INTERNATIONAL CONFERENCE ON NEURAL NETWORKS, VOLS 1-4   2   894 - 898  1997

     View Summary

    Resistive fuse is the key element for weak string filter which smoothes out noise while it detects and preserves step edges inherent to original data. Resistive fuse is implemented by two pairs of floating gate MOS transistors in a chip by a standard double poly CMOS process.

  • A Hierarchical Bayes Approach to Nonlinear Time Series Prediction with Neural Nets

    T. Matsumoto, H. Hamagishi, Y. Chonan

    ICNN'97   4   2028 - 2033  1997

  • An on-line handwriting character recognition algorithm RAV (Reparameterized Angle Variations)

    S Masaki, M Kobayashi, O Miyamoto, Y Nakagawa, T Matsumoto

    PROCEEDINGS OF THE FOURTH INTERNATIONAL CONFERENCE ON DOCUMENT ANALYSIS AND RECOGNITION, VOLS 1 AND 2   2   919 - 925  1997

     View Summary

    A new algorithm RAV (Reparameterized Angle Variations) is proposed which males explicit use of trajectory information where the time evolution of the pen coordinates plays a crucial role. The algorithm is extremely robust against stroke connections ('' Truzukeji '') as well as shape distortions ('' Kuzushi-ji ''). Preliminary experiments are reported on tests against the Kuchibese-d-96-02 data base from Tokyo University of Agriculture and Technology.

  • A fast HMM algorithm for on-line handwritten character recognition

    K Takahashi, H Yasuda, T Matsumoto

    PROCEEDINGS OF THE FOURTH INTERNATIONAL CONFERENCE ON DOCUMENT ANALYSIS AND RECOGNITION, VOLS 1 AND 2   1   369 - 375  1997

     View Summary

    A fast HMM algorithm is proposed for on-line hand written character recognition. After preprocessing input stroke are discretized so that a discrete HMM is used. This particular discretization naturally leads to a simple procedure for assigning initial state and state transition probabilities. In the training phase, complete marginelization with respect to state is not performed(Constrained Viterbi). A simple smoothing/flooring procedure yields fast and robust learning. A criterion based on normalized maximum likelihood ratio is given for deciding when to create a new model for the same character in the learning phase, in order to cope with stroke order variations and large shape variations. Preliminary experiments are done on the new Kuchibue data base from Tokyo University of Agriculture and Technology. The results seem encouraging.

  • A Hierarchical Bayesian Deconvolution with Positivity Constraints

    T.Satoh, A.Matsui, E.Okada, T.Matsumoto

    Proceedings of the 29th ISCIE International Symposium Stochastic Systems Theory and Applications     195 - 200  1997

  • Nonlinear Time Series Predictions via Hierarchical Bayes approach

    T. Matsumoto, H. Hamagishi, Y. Chonan

    Proceedings of the 29th ISCIE International Symposium Stochastic Systems Theory and Applications     239 - 244  1997

  • A Hierarchical Bayesian Regularizer Comparison for Regularization Problems

    S. Nakazawa, R. Takeuchi, T. Matsumoto

    Proceedings of the 29th ISCIE International Symposium Stochastic Systems Theory and Applications     115 - 120  1997

  • Sheet structure in global bifurcations of a driven R-L-diode circuit

    S Tanaka, SI Higuchi, T Matsumoto

    PHYSICAL REVIEW E   54 ( 6 ) 6014 - 6028  1996.12

     View Summary

    Sheet structure is found in a global bifurcation diagram of an R-L-diode circuit driven by a sinusoidal voltage source E sin2 pi ft. Bifurcations of a driven R-L-diode circuit have three interesting features: (1) The alternate appearance of large periodic windows and chaotic bands, where the period of each window increases exactly by one as E is increased. (2) The repeated appearance of period-1 attractors and chaotic bands as E is increased. (3) The existence of two different windows, each of period 2, 3, and 4. This paper attempts to provide a complete understanding of the global nature of the above features. Comprehending global bifurcations of systems, including chaotic behavior, naturally necessitates understanding the nature of stable and unstable periodic orbits, the latter being essential in most situations. The R-L-diode circuit is no exception. This paper accomplishes such a task by (i) performing extensive measurements of bifurcations in the (f,E) plane, (ii) simplifying the dynamics of the circuit without losing essential features of the observed bifurcations, and (iii) carefully analyzing the simplified dynamics from a global perspective. An analytical method in this paper is in (iii), where exact bifurcation equations are derived then the bifurcation diagrams are drawn in the (f,E,S/T) space instead of on the (f,E) plane. Here f and E are the frequency and the amplitude of the driving voltage source, and S/T will be precisely defined. This three-dimensional picture reveals thr properties of stable and unstable periodic orbits, and makes many of the global bifurcation mechanisms involved almost transparent In particular, the following are found: (1) All the period-1 attractors and their associated unstable period-1 orbits constitute a sheet structure in the (f,E,S/T) space, and hence belong to the same family. (2) Other periodic attractors of the same period and their associated unstable periodic orbits form a sheet structure in (f,E,S/T) space, and therefore belong to the same respective families. A very good correspondence between the numerical and experimental results is obtained. The global structure revealed will also clarify the global bifurcation mechanisms of other systems, e.g., the gear meshing and the offshore compliant systems described by equations similar to the present system.

    DOI PubMed CiNii

  • Recognizing chaotic time-waveforms in terms of a parametrized family of nonlinear predictors

    Tokuda, I, S Kajiwara, R Tokunaga, T Matsumoto

    PHYSICA D   95 ( 3-4 ) 380 - 395  1996.09

     View Summary

    Consider a chaotic dynamical system which exhibits a variety of chaotic time-waveforms with a change in the bifurcation parameters. This paper presents an algorithm for estimating the underlying bifurcation parameters of the chaotic time-waveforms in experimental situation in which no a priori analytical knowledge of the dynamical system is available. First. we construct ''qualitatively similar'' parametrized family of nonlinear predictors only from several sets of chaotic time-waveforms. ''Qualitatively similar'' parametrized family means that the family of nonlinear predictors exhibits ''qualitatively similar'' bifurcation phenomena as the original. Chaotic time-waveforms are then characterized in terms of the ''qualitatively similar'' bifurcation parameters of the nonlinear predictors. We call the characterization of chaotic time-waveforms in terms of the underlying bifurcation parameters ''chaotic time-waveform recognition'', Several numerical experiments using the Rossler equations show the efficiency of the algorithm. The effect of observational noise included in chaotic time-waveforms is also considered.

    DOI CiNii

  • A massively parallel resistive network for ''weak rod'': A double-layer architecture with nearest-neighbor connections

    T Matsumoto, K Kondo

    NEURAL NETWORKS   9 ( 3 ) 523 - 541  1996.04

     View Summary

    A weak rod can be realized by a very natural double-layer network with feedback. While the second layer is linear, the first layer is significantly nonlinear. Elements associated with the first layer are called ''synaptic fuses''. An MRF (Markov random field) argument is used to eliminate the line variables. The network makes transparent those portions cut when a crease or a step edge is present. Copyright (C) 1996 Elsevier Science Ltd

    DOI CiNii

  • ビジョンチップの動向

    松本隆, 小林 晴夫, 八木哲也

    光学   25 ( 5 ) 258 - 264  1996

  • ビジョンチップ-スマート画像センサ「生物の視覚情報処理と人工の目特集号」

    小林晴夫, 八木哲也, 松本 隆

    システム/制御/情報   40 ( 1 ) 13 - 18  1996

  • ダイナミカルノイズに隠された力学系とその分岐構造の推定

    手賀俊行, 徳永隆治, 梶原志保子, 松本 隆

    統計数理研究所共同研究レポート   81   12 - 25  1996

  • R-L-Diode回路のPecora-Carrol 同期:実験/理論

    西 正信, 松本 隆

    電気学会「カオス/数理と新技術」研究会   IP-96-1   1 - 10  1996

  • A Fast Convolver by Floating-Gate Transistor/Photosensor Array with Feedback

    SAKAI Takeyasu, MATSUMOTO Takashi

      ECT 96-20 ( 100 ) 75 - 80  1996

    CiNii

  • Maximum Evidence Nonlinear Time Series Prediction and Applications

    Matsumoto Takashi, Chonan Yoshimasa, Hamada Masayuki

    RIMS Kokyuroku   938   33 - 46  1996

    CiNii

  • ベイズ的標準正則化:ハイパラメータ及びRegularizerの推定法

    中沢 進, 竹内 亮, 松本 隆

    平成8年電気学会全国大会   3   36 - 37  1996

  • R-L-Diode 回路のPecora-Carrol 同期

    西 正信, 松本 隆

    平成8年電気学会全国大会   1   13  1996

  • フローティングゲートトランジスタ・アレイによる高速並列空間フィルタ

    坂井丈泰, 松本 隆

    1996 年電子情報通信学会 総合大会講演論文集   C-2   196  1996

  • フローティングゲートを用いた演算方式による高速並列空間フィルタ・チップの検討

    坂井丈泰, 松本 隆

    テレビジョン学会技術報告 IPU 96-11, CE 96-4   20 ( 23 ) 13 - 18  1996

  • 少ないデータで学習可能なHMM を用いたオンライン手書き文字認識

    高橋賢一郎, 松本 隆

    平成8年電気学会全国大会   3   42  1996

  • Bayes 法による Feedforward Neural Network における最適モデルの決定

    吉村尚郎, 松本 隆

    平成8年電気学会全国大会   3   181 - 182  1996

  • 標準正規化のハイパラメータ及びRegularizer の推定法について

    中沢 進, David J. C. Mackay, 竹内 亮, 松本 隆

    電子情報通信学会、第9回 回路とシステム軽井沢ワークショップ論文集     49 - 54  1996

  • R-L-Diode 回路のPecora-Carrol カオス同期

    西 正信, 松本 隆

    電子情報通信学会、第9回 回路とシステム軽井沢ワークショップ`論文集     177 - 182  1996

  • R-L-Diode回路のPecora-Carrol カオス同期

    松本 隆, 西 正信

    第12回ファジーシステムシンポジウム講演論文集   12   337 - 340  1996

  • ダイナミカルノイズに隠された分岐耕造の推定可能性

    手賀俊行, 徳永隆治, 松本 隆, 徳田 功

    ファジイシステムシンポジウム講演論文集   12   345 - 348  1996

  • Realization of Resistive Fuse by νMOS Transistors

    SAKAI Takeyasu, SAWAJI Toshiaki, MATSUMOTO Takashi

    The Journal of The Institute of Image Information and Television Engineers   50 ( 6 ) 783 - 786  1996

    DOI CiNii

  • Sights and Sounds of Complex Signals Generated from Simple Circuits

    T.Matsumoto

    Workshop on Generation of Digital Signals by Simple Nonlinear Devices    1996

  • An Image Compression Sensor with Analog DCT Operations

    Sakai Takeyasu, Nagai Hiromasa, Mastumoto Takashi

    Proceedings of The ITE Annual Convention   32   41 - 42  1996

     View Summary

    A parallel architecture is proposed for analog DCT operations using floating-gate transistors with differential inputs and feedback. All operations are in voltage mode. Only one terminal is required for the feedback which is capable of suppressing the distortions due to active elements. Implementing DCT operation circuits together with an array of photosensors, an image compression sensor can be realized.

    DOI CiNii

  • 非線形予測パラメータ族によるカオス的時系列データの確認

    梶原志保子, 徳田 功, 徳永隆治, 松本 隆

    電子情報通信学会論文誌A   J79- A ( 8 ) 1394 - 1403  1996

  • システムノイズを含むカオス的時系列のMaximum Evidence 予測

    浜岸広明, 松本 隆

    日本神経回路学会、第7回全国大会     110 - 111  1996

    CiNii

  • オンライン手書き文字認識アルゴリズムR.A.V. 認識率を向上させるストローク位置対応の考慮

    真崎晋哉, 中川洋一, 松本 隆

    電子情報通信学会1996年情報・システムソサイエティ大会講演論文集   D-367   370  1996

  • A Bayesian Nonlinear Time Series Prediction

    MATSUMOTO T.

    Proc. World Congress on Neural Networks, DanDiego     659 - 667  1996

    CiNii

  • 画像圧縮イメージセンサのためのアナログDCT演算回路の検討

    坂井丈泰, 永井宏昌, 松本 隆

    1996 年電子情報通信学会エレクトロニクスソサイエティ大会講演文集   C-531 ( 2 ) 194  1996

  • フローティングゲート付差動増幅器のためのリセット回路について

    坂井丈泰, 永井宏昌, 松本 隆

    テレビジョン学会技術報告   20 ( 55 ) 43 - 48  1996

  • 標準正則化のハイパーパラメータおよびRegularizerの推定法について

    竹内亮, David J. C. MacKay, 中沢進, 松本隆

    電子情報通信学会論文誌D-II   J79-D-II ( 12 ) 1 - 9  1996

  • Maximum Evidence Nonlinear Time Series Prediction and Applications

    T.Matsumoto, Y.Chonan, M.Hamada

    京都大学数理解析研究所講究録 938、低次元力学系とその周辺   938   33 - 46  1996

    CiNii

  • Sights and Sounds of Complex Signals Generated from Simple Circuits

    T.Matsumoto

    Workshop on Generation of Digital Signals by Simple Nonlinear Devices    1996

  • A Bayesian Nonlinear Time Series Prediction

    T.Matsumoto, Y.Chonan, M.Hamada

    WCNN International Neural Network Society Annual Meeting     659 - 667  1996

  • Vision Chip Architecture with Light Adaptation Mechanism

    H.Kobayashi, T.Matsumoto, T.Yagi, K.Tanaka

    Proc. International Symposium on Artificial Life and Robotics (AROB)    1995

  • 陰影からの3次元形状復元に関する一考察

    坂本 聡生, 松本 隆, 藤井真人, 伊藤 崇之

    1995年電子情報通信学会情報・システムソサイエティ大会講演論文集   D-222   225  1995

  • 標準正則化問題の最適ハイパーパラメータ及びregularizerの決定法について

    竹内亮, D.J. C. MacKay, 松本 隆

    電子情報通信学会技術研究報告(NC-94-60)   94 ( 487 ) 1 - 8  1995

  • Reparametrized Angle Variationを用いるon-line手書き文字認識

    小林充, 宮本修, 森哲也, 中川洋一, 松本 隆

    電子情報通信学会技術研究報告   94 ( 509 ) 23 - 30  1995

  • Realization of the Weak Rod Filter by a Double Layer Network

    SAKAI Takeyasu, KONDO Kenji, MATSUMOTO Takashi

    ITEJ Technical Report   19 ( 7 ) 25 - 30  1995

     View Summary

    The weak string filter smoothes out noise involved in a given data while it detects step edges inherent to the original noise free data. However, a steep ramp in intensity often cannot be preserved by this filter. The weak rod filter which contains the second spatial difference in its regularizer, solves this problem, and can be realized by a double layer resistive network.

    CiNii

  • ニューラルネットへのARD (AutomaticRelevance Determination) 適用の有効性について

    高木健次, 松本 隆

    電子情報通信学会技術研究報告(NC-94-61)   94 ( 487 ) 9 - 16  1995

  • 明るさ知覚モデル6層回路網

    村橋英樹, 松本 隆

    電子情報通信学会総合大会   D-46   51  1995

  • 6層構造を持つ明るさ知覚モデル回路網

    村橋英樹, 松本 隆

    電子情報通信学会技術研究報告(NC-94-88)   94 ( 562 ) 95 - 102  1995

  • ラインプロセスを持つフィルタのアナログ回路による実現

    坂井丈泰, 沢地利明, 近藤堅司, 松本 隆

    電子情報通信学会技術研究報告(NC-94-135)   94 ( 563 ) 163 - 170  1995

  • On-line文字認識アルゴリズムReparametrized Angle Variationsを高速に実行するハードウェアボードについて

    宮本 修, 中川洋一, 松本 隆

    電子情報通信学会技術研究報告   94 ( 548 ) 49 - 56  1995

  • Light Adaptive Parallel Analog Image Processors(光適応並列アナログ画像処理プロセッサ)

    小林春夫, 松本 隆, 八木哲也, 田中孝治

    テレビジョン学会技術報告   19 ( 25 ) 7 - 12  1995

  • 多重ハイパーパラメータ正則化モデルとその応用について

    松井 淳, 松本 隆

    電子情報通信学会技術研究報告(NC95-38)   95 ( 189 ) 71 - 78  1995

  • Maximum Evidence時系列予測について

    松本 隆, 長南吉正, 浜田雅之

    電気学会研究会 情報処理研究会   IP-95-31   1 - 10  1995

  • Two-Dimentional Spatio-Temporal Dynamics of Analog Image Processing Neural Networks

    H.Kobayashi, T.Matsumoto, J.Sanekata

    IEEE Transactions on Neural Networks   6 ( 5 ) 1148 - 1164  1995

    DOI CiNii

  • カオス的時間連続力学系に対する分岐図再構成

    梶原志穂子, 徳田功, 徳永隆治, 松本 隆

    電気学会研究会 情報処理研究会   IP-95-32 ( 31 ) 11 - 21  1995

    CiNii

  • Feedforward Neural NetによるMaximum Evidence時系列予測

    長南吉正, 浜田雅之, 松本 隆

    日本神経回路学会 第6回全国大会講演論文集   P2-29   195 - 196  1995

  • SCE Vision Chip

    松本 隆, 小林春夫, 八木哲也

    計測自動制御学会,第15回光応用計測部会講演会   95PG0010   7 - 16  1995

  • ベイズ的手法による正則化

    松井 淳, 中沢 進, 竹内 亮, 松本 隆

    日本神経回路学会 第6回全国大会講演論文集   P1-6   60 - 63  1995

  • ラインプロセスを含むフィルタのアナログCMOS回路設計

    坂井丈泰, 松本 隆

    日本神経回路学会 第6回全国大会講演論文集   P2-39   213 - 214  1995

  • “Evidence”によるFeedforward Neural Networkの評価

    吉村尚郎, 松本 隆

    日本神経回路学会 第6回全国大会講演論文集   P1-5   58 - 59  1995

  • ビジョンチップの最新動向

    松本隆

    第7回画像入力シンポジウム予稿集    1995

  • 多重weight decayニューラルネットワークによる学習

    松本晴幸, 土屋和広, 山本正典, 黒谷憲一, 若原邦夫, 吉村尚郎, 高木健次, 松本 隆

    自動制御連合講演会    1995

  • Vision Chip Architecture with Light Adaptation Mechanism

    H.Kobayashi, T.Matsumoto, T.Yagi, K.Tanaka

    Proc. International Symposium on Artificial Life and Robotics (AROB)    1995

  • LIGHT-ADAPTIVE ARCHITECTURES FOR REGULARIZATION VISION CHIPS

    H KOBAYASHI, T MATSUMOTO, T YAGI, K TANAKA

    NEURAL NETWORKS   8 ( 1 ) 87 - 101  1995

     View Summary

    Light-adaptive algorithms/architectures are proposed for regularization vision chips. The adaptation mechanisms allow the regularization parameters to change in an adaptive manner in accordance with the light intensity of given images. This is achieved by adaptively changing the conductance values associated with massively parallel resistive networks. The algorithms/architectures are inspired by the adaptation mechanisms of the horizontal cells in the lower vertebrate retina.

    DOI CiNii

  • Two-Dimensional Spatio-Temporal Dynamics of Analog Image Processing Neural Networks

    Haruo Kobayashi, Takashi Matsumoto, Jun Sanekata

    IEEE Transactions on Neural Networks   6 ( 5 ) 1148 - 1164  1995

     View Summary

    A typical analog image-processing neural network consists of a two-dimensional (2-D) array of simple processing elements. When it is implemented with CMOS LSI, two dynamics issues naturally arise: 1) Parasitic capacitors of MOS transistors induce temporal dynamics. Since a processed image is given as the stable equilibrium point of temporal dynamics, a temporally unstable chip is unusable. 2) Because of the array structure, the node voltage distribution induces spatial dynamics, and the node voltage distribution could behave in a wild manner, e.g., oscillatory, which is undesirable for image-processing purposes. A discussion of these issues for one-dimensional cases is found in [1]. This paper extends its results to 2-D cases and also derives several explicit formulas and relationships for the 2-D dynamics, which are useful for the design and analysis of the class of networks of interest. Specifically, the following are derived: i) explicit spatial and temporal stability conditions and their equivalency, ii) spatial impulse responses, iii) spatial frequency responses, iv) power consumption, v) time constants, vi) relationships between spatial frequency responses and stability, vii) relationships between power consumption and stability, viii) relationships between spatial impulse responses and the discrete Fourier transform of network parameters, ix) relationships between spatial impulse responses and the inverse Z-transform of a transfer function, x) relationships between spatial frequency responses and time constants, xi) relationships between spatial frequency responses and equivalent circuits, xii) the characteristics of stable and unstable network dynamics, and xiii) hexagonal as well as square grid network dynamics. © 1995 IEEE

    DOI CiNii

  • REALIZATION OF THE WEAK ROD BY A DOUBLE-LAYER PARALLEL NETWORK

    T MATSUMOTO, K KONDO

    NEURAL COMPUTATION   6 ( 5 ) 944 - 956  1994.09

     View Summary

    The weak rod can be realized by a very natural double layer network with feedback, While the second layer is linear, the first layer is significantly nonlinear. Elements associated with the first layer are called ''synaptic fuses.'' An MRF argument is used to eliminate the line variables, The network makes what is cut transparent when a crease or a step edge is present.

  • 順応機能をもつビジョンチップ

    T.Yagi, T.Matsumoto, H.Kobayashi, K.Tanaka

    日本機械学会第3回バイオエンジニアリングシンポジウム講演論文集   940 ( 5 ) 78 - 79  1994

  • Reconstructing bifurcation diagrams only from time-waveforms

    R.Tokunaga, S.Kajiwara, T.Matsumoto

    Physica D   79 ( 2/4 ) 348 - 360  1994

    DOI CiNii

  • 視覚前注意過程におけるテクスチャー境界検出モデル

    佐野雅規, 伊藤崇之, 中川俊夫, 松本隆

    1994年電子情報通信学会春季大会講演論文集   6   77  1994

  • A model of preattentive texture boundary detection

    SANO Masanori, ITO Takayuki, NAKAGAWA Toshio, MATSUMOTO Takashi

    ITEJ Technical Report   18 ( 24 ) 31 - 36  1994

     View Summary

    This paper presents a model of preattentive texture boundary detection. This model is constracted based on physiological and psychological data. The model consists of three stages. In the first stage Gabor filter decomposition provides representations of an image : orientation selectivity, spatial frequency. Following this, five modules for feature extraction are introduced : brightness contrast, scale contrast, orientation contrast, hyper-orientation contrast and end-of-line density contranst. Each module outputs high value where its feature's value change quickly among neighbouring area. In the final stage these information are compared and intergrared into the final output. The model has been tested with several classical stimuli from psychophysical literature. According to the results of experiments, this model has an ability to detect texture boundaries where we perceive ones in preattentive process.

    CiNii

  • Spatial and Temporal Dynamics of Analog Image-Processing Neuro Chips

    Kobayashi Haruo, Matsumoto Takashi, Sanekata Jun

    IEICE technical report. Computer systems   94 ( 15 ) 31 - 38  1994

     View Summary

    There are two dynamics issues in analog image processing neuro chips: (i)The temporal dynamics issue due to the paxasitic capacitors in a CMOS chip,and (ii)The spatial dynamics issue due to the regulax array of processing elements in a chip. This paper extends our previous results to 2D case and sharpens them in the following sense: (i)2D spatial stability definition which is consistent with the previous 1D case is given. (ii)Explicit spatial and temporal stability conditions are given. (iii)Explicit spatial frequency and impulse responses are given. (iv)An explicit formula for time constants is given. (v)Several relationships between the spatial and temporal dynamics are given.

    CiNii

  • Reconstructing Bifurcation Diagrams only from Time-Waveforms

    R.Tokunaga, I.Tokuda, S.Kajiwara, T.Matsumoto

    Proceedings of the International Conference on Dynamical Systems and Chaos   2   345 - 354  1994

  • 時系列データからの分岐図再構成

    梶原志穂子, 徳永隆治, 松本隆

    電子情報通信学会論文誌A   J77-A ( 3 ) 408 - 419  1994

  • 分岐について

    松本 隆

    数理科学、6月号   32 ( 6 ) 5 - 8  1994

  • A Piecewise-Linear Regression on the ASHRAE Time- Series Data

    M.Iijima, K.Takagi, R.Takeuchi, T.Matsumoto

    ASHRAE TRANSACTIONS OR-94-17-5   100 ( 2 )  1994

  • Reconstructing Bifurcation Diagrams only from Chaotic Time-Wave Forms

    S.Kajiwara, I.Tokuda, R.Toknaga, T.Matsumoto

    3rd International Conference on Fuzzy Logic, Neural Nets and Soft Computing     515 - 517  1994

  • 標準正則化における正則化パラメータの一決定法

    竹内 亮, 松本 隆

    1994年電子情報通信学会秋季大会講演論文集,A-136     136  1994

  • ニューラルネットワークにおけるARDについて

    高木健次, 松本 隆

    1994年電子情報通信学会秋季大会     32  1994

  • R-L-Diode回路のカオス的同期

    栗本康行, 松本 隆

    1994年電子情報通信学会秋季大会     44  1994

  • Bifurcations of the Double Scroll Circuit

    T.Matsumoto, R.Tokunaga, M.Komuro

    Nonlinearity and Chaos in Engineering Dynamics    1994

  • カオス的同期現象について

    栗本康行, 松本 隆

    マイクロエレクトロニクス研究開発機構第11回研究交流会    1994

  • 区分線形ベクトル場におけるOrbit-flip Homoclinic Orbitの分岐

    栗本康行, 庵勝仁, 松本隆

    電気学会、情報処理研究会   IP94 ( 14 ) 1 - 10  1994

  • ビジョンチップI,II:アナログ画像処理用ニューロチップ

    小林春夫, 松本 隆, 八木哲也

    システム制御情報学会、応用信号処理分科会、第4回研究会     9 - 26  1994

  • The Weak String Filter and Its Implementation with vMOS Transistors

    坂井丈泰, 澤地利明, 松本 隆

    テレビジョン学会技術報告、情報入力、情報ディスプレイ合同研究会   18 ( 68 ) 13 - 18  1994

  • ARD(Automatic Relevance Determination)のM.L.P.への適用について

    高木健次, 松本 隆

    日本神経回路学会、第5回全国大会     305 - 306  1994

  • Reality of Chaos in the Double Scroll Circuit – A Computer-Assisted Proof

    T. Matsumoto, T. Ayaki

    IEEE Trans. On Circuits & Systems I – Fundamental Theory and Applications   41 ( 11 ) 736 - 740  1994

  • Determining Optimal Hyperparameters and Regularizers for Standard Regularization Problems

    R. Takeuchi, David, J. Mackay, T. Matsumoto

    1994 International Symposium on Artificial Neural Networks Proceedings     419 - 428  1994

  • Reconstructing bifurcation diagrams only from time-waveforms

    R.Tokunaga, S.Kajiwara, T.Matsumoto

    Physica D   79 ( 2/4 ) 348 - 360  1994

    DOI CiNii

  • Reconstructing Bifurcation Diagrams only from Time-Waveforms

    R.Tokunaga, I.Tokuda, S.Kajiwara, T.Matsumoto

    Proceedings of the International Conference on Dynamical Systems and Chaos   2   345 - 354  1994

  • A Piecewise-Linear Regression on the ASHRAE Time- Series Data

    M.Iijima, K.Takagi, R.Takeuchi, T.Matsumoto

    ASHRAE TRANSACTIONS OR-94-17-5   100 ( 2 )  1994

  • Reconstructing Bifurcation Diagrams only from Chaotic Time-Wave Forms

    S.Kajiwara, I.Tokuda, R.Toknaga, T.Matsumoto

    3rd International Conference on Fuzzy Logic, Neural Nets and Soft Computing     515 - 517  1994

  • Bifurcations of the Double Scroll Circuit

    T.Matsumoto, R.Tokunaga, M.Komuro

    Nonlinearity and Chaos in Engineering Dynamics    1994

  • Reality of Chaos in the Double Scroll Circuit – A Computer-Assisted Proof

    T. Matsumoto, T. Ayaki

    IEEE Trans. On Circuits & Systems I – Fundamental Theory and Applications   41 ( 11 ) 736 - 740  1994

  • Determining Optimal Hyperparameters and Regularizers for Standard Regularization Problems

    R. Takeuchi, David, J. Mackay, T. Matsumoto

    1994 International Symposium on Artificial Neural Networks Proceedings     419 - 428  1994

  • 連続区分線形電子回路におけるカオスの計算機援用証明

    徳永隆治, 松本 隆

    日本応用数理学会,平成5年度年会講演予稿集     21 - 22  1993

  • Observing a Codimension-Two Heteroclinic Bifurcation

    R.Tokunaga, Y.Abe, T.Matsumoto

    Chaos – An Interdisciplinary Journal of Nonlinear Science   3 ( 1 ) 63 - 72  1993

  • Early Vision Chips

    T.Matsumoto

    1993年電気情報関連学会連合大会講演論文集     125 - 136  1993

  • ビジョンチップI

    松本 隆, 小林春夫, 八木哲也

    電子情報通信学会雑誌   76 ( 7 ) 783 - 791  1993

  • ビジョンチップII

    松本 隆, 小林春夫, 八木哲也

    電子情報通信学会雑誌   76 ( 8 ) 851 - 858  1993

  • ビジョンチップ

    小林春夫, 松本 隆, 八木哲也

    精密工学会,画像応用技術専門委員会研究会報告   8 ( 2 ) 9 - 26  1993

  • 視覚センサー(ビジョンチップ)

    小林春夫, 松本 隆, 八木哲也

    第5回画像入力シンポジウム    1993

  • Observing a Codimension-Two Heteroclinic Bifurcation

    R.Tokunaga, Y.Abe, T.Matsumoto

    Chaos – An Interdisciplinary Journal of Nonlinear Science   3 ( 1 ) 63 - 72  1993

  • WEAK ROD BY A DOUBLE-LAYER PARALLEL NETWORK

    T MATSUMOTO, K KONDO

    IJCNN '93-NAGOYA : PROCEEDINGS OF 1993 INTERNATIONAL JOINT CONFERENCE ON NEURAL NETWORKS, VOLS 1-3   1   857 - 864  1993

  • SPATIAL AND TEMPORAL DYNAMICS OF ANALOG IMAGE-PROCESSING NEURO CHIPS

    H KOBAYASHI, T MATSUMOTO, J SANEKATA

    IJCNN '93-NAGOYA : PROCEEDINGS OF 1993 INTERNATIONAL JOINT CONFERENCE ON NEURAL NETWORKS, VOLS 1-3   1   865 - 874  1993

  • IMAGE-PROCESSING REGULARIZATION FILTERS ON LAYERED ARCHITECTURE

    H KOBAYASHI, T MATSUMOTO, T YAGI, T SHIMMI

    NEURAL NETWORKS   6 ( 3 ) 327 - 350  1993

     View Summary

    Layered architecture is proposed for solving a class of regularization problems in image processing. There are two major hurdles in the implementation of regularization filters with second or higher order smoothness constraints: (a) Stability: With second or higher order constraints, a direct implementation of a regularization filter necessitates negative conductance which, in turn, gives rise to stability problems. (b) Wiring Complexity: A direct implementation of an N-th order regularization filler requires wiring between every pair of k-th nearest nodes for all k, 1 less-than-or-equal-to k less-than-or-equal-to N. Even though one of the authors managed to layout an N = 2 chip, the implementation of an N greater-than-or-equal-to 3 chip would be an extremely difficult, if not impossible, task. The regularization filter architecture proposed here (a) requires no negative conductance,- and (b) necessitates wiring only between nearest nodes. Smoothing-Contrast-Enhancement filter is given as an example of application. Since this filter is extremely fast, it will have a natural application to smart sensing, i. e., to the simultaneous achievement of sensing and processing. It is also explained how this architecture has been inspired by physiological findings on lower vertebrate retina by one of the authors.

    DOI CiNii

  • A CNN HANDWRITTEN CHARACTER RECOGNIZER

    H SUZUKI, T MATSUMOTO, LO CHUA

    INTERNATIONAL JOURNAL OF CIRCUIT THEORY AND APPLICATIONS   20 ( 5 ) 601 - 612  1992.09

     View Summary

    CNNs are used for feature detection in handwritten character recognition. Detected features are fed to a simple classifier network. Performance was tested by using two well-known ETL data base series: (i) ETL3 consisting of numerals, alphabets and several symbols and (ii) ETL8B2 consisting of Japanese Hirakana characters. The average recognition rate for ETL3 is 94.8%, while that for ETL8B2 is 85.7%. Both series include 'hard' characters so distorted that even humans cannot recognize them.

    DOI CiNii

  • SPATIAL VERSUS TEMPORAL STABILITY ISSUES IN IMAGE-PROCESSING NEUROCHIPS

    T MATSUMOTO, H KOBAYASHI, Y TOGAWA

    IEEE TRANSACTIONS ON NEURAL NETWORKS   3 ( 4 ) 540 - 569  1992.07

     View Summary

    A typical image processing neuro chip consists of a regular array of very simple cell circuits. When it is implemented by a CMOS process, two stability issues naturally arise: i) Parasitic capacitors of MOS transistors induce the temporal dynamics. Since a processed image is given as the stable limit point of the temporal dynamics, a temporally unstable chip is unusable. ii) Because of the array structure, the node voltage distribution induces the spatial dynamics, and it could behave in a wild manner, e.g., oscillatory, which is highly undesirable for image processing purposes, even if the trajectory of the temporal dynamics converges to a stable limit point.
    The main contributions of this paper are (i) a clarification of the spatial stability issue; (ii) explicit if and only if conditions for the temporal and the spatial stability in terms of circuit parameters; (iii) a rigorous explanation of the fact that even though the spatial stability is stronger than the temporal stability, the set of parameter values for which the two stability issues disagree is of (Lebesgue) measure zero; and (iv) theoretical estimates on the processing speed.

    DOI CiNii

  • Codim-2 Heteroclinic 分岐の観測:実験と検証

    阿部 悌, 徳永 隆治, 松本 隆

    電子情報通信学会技術研究報告CAS91-146   NLP91-89 ( 452 ) 43 - 50  1992

  • 網膜神経回路は標準正則化問題を解く

    T.Yagi, T.Matsumoto, Y.Funabashi

    電子情報通信学会技術研究報告(NC),NC91-100   91 ( 529 ) 13 - 20  1992

  • 多重解像度表現によるテクスチャ境界の抽出

    M.Sano, T.Ito, T.Nakagawa, T.Matsumoto

    電子情報通信学会技術研究報告NC91-110   91 ( 529 ) 89 - 96  1992

  • 見えない不動点を追う:分岐トポグラフィ

    松本隆, 徳永隆治

    情報処理学会論文誌   33 ( 4 ) 384 - 399  1992

  • Observing Codim - 2 Heteroclinic Bifurcations

    R.Tokunaga, Y.Abe, T.Matsumoto

    Proc. 1992 Symp. Nonlinear Theory and its Applications (NOLTA)     107 - 110  1992

  • N-Homoclinic Bifurcations of Piecewise Linear Vector Fields

    K. Iori, E. Yanagida, T. Matsumoto

    京都大学数理解析研究所講究録   804   75 - 90  1992

  • A parallel analog CMOS signal processor for image contrast enhancement

    SHIMMI T.

    Proc. European Solid-State Circuits Conf., Copenhagen, Denmark     163 - 166  1992

    CiNii

  • A Massively Parallel CMOS Vision Chip : Double-layer Regularization Filter

    Matsumoto T., Abidi A.A., Shimmi T., Kobayashi H., Yagi T., Sawaji T.

    ITE Technical Report   16 ( 79 ) 13 - 18  1992

     View Summary

    This paper is an implementation version of our algorithm proposed in [2]. The chip implemented solves first and second order regularization problems simultaneously which, in turn, enhances contrasts of images after smoothing. A 2μm standard CMOS technology was used with double metal and single poly. The computation is done by the dynamics and its execution time is within several micro seconds.

    DOI CiNii

  • A Second Order Regularization Vision Chip for Smoothing-Contrast Enhancement

    T.Matsumoto, T.Shinmmi, H.Kobayashi, A. A. Abidi, T.Yagi, T.Sawaji

    Proc. IJCNN 1992   3   188 - 197  1992

  • Observing Codim - 2 Heteroclinic Bifurcations

    R.Tokunaga, Y.Abe, T.Matsumoto

    Proc. 1992 Symp. Nonlinear Theory and its Applications (NOLTA)     107 - 110  1992

  • Spatial Versus Temporal Stability Issues in Image Processing Neuro Chips

    Takashi Matsumoto, Haruo Kobayashi, Yoshio Togawa

    IEEE Transactions on Neural Networks   3 ( 4 ) 540 - 569  1992

     View Summary

    A typical image processing neuro chip consists of a regular array of very simple cell circuits. When it is implemented by a CMOS process, two stability issues naturally arise: i) Parasitic capacitors of MOS transistors induce the temporal dynamics. Since a processed image is given as the stable limit point of the temporal dynamics, a temporally unstable chip is unusable. ii) Because of the array structure, the node voltage distribution induces the spatial dynamics, and it could behave in a wild manner, e.g., oscillatory, which is highly undesirable for image processing purposes, even if the trajectory of the temporal dynamics converges to a stable limit point. The main contributions of this paper are (i) a clarification of the spatial stability issue
    (ii) explicit if and only if conditions for the temporal and the spatial stability in terms of circuit parameters
    (iii) a rigorous explanation of the fact that even though the spatial stability is stronger than the temporal stability, the set of parameter values for which the two stability issues disagree is of (Lebesgue) measure zero
    and (iv) theoretical estimates on the processing speed. © 1992 IEEE

    DOI CiNii

  • N-Homoclinic Bifurcations of Piecewise Linear Vector Fields

    K. Iori, E. Yanagida, T. Matsumoto

    京都大学数理解析研究所講究録   804   75 - 90  1992

  • A Parallel Analog CMOS Signal Processor for Image Contrast Enhancement

    T.Shimmi, H.Kobayashi, T.Yagi, T.Sawaji, T.Matsumoto, A.A. Abidi

    Proc. European Solid State Circuits Conference(ESSCIRC’92)     163 - 166  1992

  • A Second Order Regularization Vision Chip for Smoothing-Contrast Enhancement

    T.Matsumoto, T.Shinmmi, H.Kobayashi, A. A. Abidi, T.Yagi, T.Sawaji

    Proc. IJCNN 1992   3   188 - 197  1992

  • MULTI-FOLDING - ALTERNATIVE APPEARANCE OF PERIOD-ONE ATTRACTORS AND CHAOTIC ATTRACTORS IN A DRIVEN R-L-DIODE CIRCUIT

    S TANAKA, T MATSUMOTO, J NOGUCHI, LO CHUA

    PHYSICS LETTERS A   157 ( 1 ) 37 - 43  1991.07

     View Summary

    The paper reports multi-folding, a chaotic attractor formation mechanism which is responsible for alternative appearance of period-one attractors and chaotic attractors in a bifurcation diagram of a driven R-L-diode circuit. The multi-folding, when couched in terms of a simplified one-dimensional map is characterized by its multi-modality.

    DOI CiNii

  • GLOBAL BIFURCATION ANALYSIS OF THE DOUBLE SCROLL CIRCUIT

    M. Komuro, R. Tokunaga, T. Matsumoto, L. O. Chua, A. Hotta

    INTERNATIONAL JOURNAL OF BIFURCATION AND CHAOS   1 ( 1 ) 139 - 182  1991.03

     View Summary

    An in-depth analysis is made of the global 2-parameter bifurcation structures of the double scroll circuit in terms of their homoclinic, heteroclinic, and periodic orbits. Many fine details are uncovered via a 3-dimensional "unfolding" of the 2-parameter bifurcation structures.
    Major findings are:
    (i) The parameter sets which give rise to the homoclinic and heteroclinic orbits (homoclinic and heteroclinic bifurcation sets) studied in this paper are found to be all connected to each other via only one family of periodic orbits.
    (ii) Moreover, the structure of the windows of this family essentially determines the global structure of the periodic windows of the double scroll circuit.
    These bifurcation analyses are accomplished by deriving first the relevant bifurcation equations in exact analytic form and then solving these nonlinear equations by iterations. No numerical integration formula for differential equations are used.

    DOI

  • Linear vs. Nonlinear Method

    T.Matsumoto, Y.Togawa

    Testing and Diagnosis of Analog Circuits and Systems     37 - 63  1991

  • A Repeated Appearance of Period -1 Attractor in a Driven R-L-Diode Circuit: Experimental and Theoretical Bifurcation Analysis

    S.Tanaka, J.Noguchi, S.Higuchi, T.Matsumoto

    IEICE Trans.   E74 ( 6 ) 1406 - 1413  1991

  • 精度保証付き算法を用いたカオスの厳密証明

    徳永隆治, 松本隆

    電子情報通信学会技術研究報告   NLP91 ( 4 ) 1 - 8  1991

  • Stability of Image Processing Neuro Chips: Spatial and Temporal

    T.Matsumoto, H. Kobayashi, Y. Togawa

    Proc. IJCNN   2   283 - 296  1991

  • Non-Symmetric CNN Templates for Image Processing

    T.Yokohama, H.Suzuki, Y.Matsushita, T.Matsumoto, L. O. Chua

    European Conference on Circuit Theory and Design-91     10 - 19  1991

  • A Layered Architecture for Regularization Vision Chips

    H.Kobayashi, T.Matsumoto, T.Yagi, T.Shimmi

    Proc.IJCNN   2 ( 3 ) 1007 - 1019  1991

  • Linear vs. Nonlinear Method

    T.Matsumoto, Y.Togawa

    Testing and Diagnosis of Analog Circuits and Systems     37 - 63  1991

  • A Repeated Appearance of Period -1 Attractor in a Driven R-L-Diode Circuit: Experimental and Theoretical Bifurcation Analysis

    S.Tanaka, J.Noguchi, S.Higuchi, T.Matsumoto

    IEICE Trans.   E74 ( 6 ) 1406 - 1413  1991

  • Stability of Image Processing Neuro Chips: Spatial and Temporal

    T.Matsumoto, H. Kobayashi, Y. Togawa

    Proc. IJCNN   2   283 - 296  1991

  • Non-Symmetric CNN Templates for Image Processing

    T.Yokohama, H.Suzuki, Y.Matsushita, T.Matsumoto, L. O. Chua

    European Conference on Circuit Theory and Design-91     10 - 19  1991

  • A Layered Architecture for Regularization Vision Chips

    H.Kobayashi, T.Matsumoto, T.Yagi, T.Shimmi

    Proc.IJCNN   2 ( 3 ) 1007 - 1019  1991

  • Bifurcation Analysis of Shilnikov's Chaos

    R.Fujimoto, A.Hotta, R.Tokunaga, M.Komuro, T.Matsumoto

    World Scientific Advanced Series in Dynamical Systems   8   125 - 142  1990

  • Image Thinning With a Cellular Neural Network

    T.Matsumoto, L. O. Chua, T.Yokohama

    IEEE Trans. CAS   37 ( 5 ) 638 - 640  1990

    DOI CiNii

  • CNN Cloning Template: Hole-Filler

    T. Matsumoto, L. O. Chua, R. Furukawa

    IEEE Trans. CAS   37 ( 5 ) 635 - 638  1990

    DOI CiNii

  • CNN Cloning Template: Connected Component Detector

    T. Matsumoto, L. O. Chua, H. Suzuki

    IEEE Trans. CAS   37 ( 5 ) 633 - 635  1990

    DOI CiNii

  • The Piecewise-Linear Lorenz Circuit is Chaotic in the Sense of Shilnikov

    R. Tokunaga, T. Matsumoto, L. O. Chua, S. Miyama

    IEEE Trans. CAS   37 ( 6 ) 766 - 786  1990

    DOI CiNii

  • A Hierarchical Structure in Homoclinic Bifurcations

    R. Fujimoto, M. Komuro, R. Tokunaga, T. Matsumoto

    Trans. IEICE   E73 ( 6 ) 809 - 816  1990

  • CNN Cloning Template: Shadow Detector

    T. Matsumoto, L. O. Chua, H. Suzuki

    IEEE TRANS. CAS   37 ( 8 ) 1070 - 1073  1990

    DOI CiNii

  • Several Image Processing Examples by CNN

    T.Matsumoto, T.Yokohama, H.Suzuki, R.Furukawa, A.Oshimoto, T.Shimmi, Y.Matsushita, T.Seo, L. O. Chua

    IEEE First International Workshop on Cellular Neural Networks and Their Applications   CNNA-90   100 - 111  1990

  • Bifurcation Analysis of Shilnikov's Chaos

    R.Fujimoto, A.Hotta, R.Tokunaga, M.Komuro, T.Matsumoto

    World Scientific Advanced Series in Dynamical Systems   8   125 - 142  1990

  • Image Thinning With a Cellular Neural Network

    T.Matsumoto, L. O. Chua, T.Yokohama

    IEEE Trans. CAS   37 ( 5 ) 638 - 640  1990

    DOI CiNii

  • CNN Cloning Template: Hole-Filler

    T. Matsumoto, L. O. Chua, R. Furukawa

    IEEE Trans. CAS   37 ( 5 ) 635 - 638  1990

    DOI CiNii

  • CNN Cloning Template: Connected Component Detector

    T. Matsumoto, L. O. Chua, H. Suzuki

    IEEE Trans. CAS   37 ( 5 ) 633 - 635  1990

    DOI CiNii

  • The Piecewise-Linear Lorenz Circuit is Chaotic in the Sense of Shilnikov

    R. Tokunaga, T. Matsumoto, L. O. Chua, S. Miyama

    IEEE Trans. CAS   37 ( 6 ) 766 - 786  1990

    DOI CiNii

  • A Hierarchical Structure in Homoclinic Bifurcations

    R. Fujimoto, M. Komuro, R. Tokunaga, T. Matsumoto

    Trans. IEICE   E73 ( 6 ) 809 - 816  1990

  • CNN Cloning Template: Shadow Detector

    T. Matsumoto, L. O. Chua, H. Suzuki

    IEEE TRANS. CAS   37 ( 8 ) 1070 - 1073  1990

    DOI CiNii

  • Several Image Processing Examples by CNN

    T.Matsumoto, T.Yokohama, H.Suzuki, R.Furukawa, A.Oshimoto, T.Shimmi, Y.Matsushita, T.Seo, L. O. Chua

    IEEE First International Workshop on Cellular Neural Networks and Their Applications   CNNA-90   100 - 111  1990

  • Homoclinic Linkage: A New Bifurcation Mechanism

    R.Tokunaga, T.Matsumoto, M.Komurao, L. O. Chua, K.Miya, A.Hotta, R.Fujimoto

    Proc. ISCAS   2   826 - 829  1989

  • Homoclinic Linkage in the Double Scroll Circuit and the Cusp-Constrained Circuit

    R.Tokunaga, T.Matsumoto, T.Ida, K. Miya

    World Scientific Advanced Series in Dynamical Systems, The Study of Dynamical Systems   7   192 - 209  1989

  • Homoclinic Linkage in the Double Scroll Circuit and the Cusp-Constrained Circuit

    R.Tokunaga, T.Matsumoto, T.Ida, K.Miya

    京都大学数理解析研究所講究録   696   204 - 221  1989

  • Bifurcation Analysis of Shilnikov's Chaos

    R.Fujimoto, A.Hotta, R.Tokunaga, M.Komuro, T.Matsumoto

    京都大学数理解析研究所講究録   710   127 - 144  1989

  • Homoclinic分岐の階層構造

    藤本竜一, 小室元政, 徳永隆治, 松本隆

    電子情報通信学会技術研究報告   NLP89 ( 61 ) 59 - 64  1989

  • カオスを聴く

    松本隆, 小室元政, L.O.Chua

    Computer Today   6 ( 32 ) 40 - 42  1989

  • Homoclinic Linkages:Double Scroll回路

    宮和行, 徳永隆治, 堀田篤宏, 藤本竜一, 松本隆, 小室元政, L.O.Chua

    電子情報通信学会技術研究報告   NLP88 ( 68 ) 9 - 16  1989

  • Bifurcation Analysis of A Cusp-Constrained Piecewise-linear Circuit

    R.Tokunaga, L. O. Chua, T.Matsumoto

    Int. J.CTA   17 ( 3 ) 283 - 346  1989

    DOI CiNii

  • Homoclinic Linkage: A New Bifurcation Mechanism

    R.Tokunaga, T.Matsumoto, M.Komurao, L. O. Chua, K.Miya, A.Hotta, R.Fujimoto

    Proc. ISCAS   2   826 - 829  1989

  • Homoclinic Linkage in the Double Scroll Circuit and the Cusp-Constrained Circuit

    R.Tokunaga, T.Matsumoto, T.Ida, K. Miya

    World Scientific Advanced Series in Dynamical Systems, The Study of Dynamical Systems   7   192 - 209  1989

  • Homoclinic Linkage in the Double Scroll Circuit and the Cusp-Constrained Circuit

    R.Tokunaga, T.Matsumoto, T.Ida, K.Miya

    京都大学数理解析研究所講究録   696   204 - 221  1989

  • Bifurcation Analysis of Shilnikov's Chaos

    R.Fujimoto, A.Hotta, R.Tokunaga, M.Komuro, T.Matsumoto

    京都大学数理解析研究所講究録   710   127 - 144  1989

  • LORENZ ATTRACTOR FROM AN ELECTRICAL CIRCUIT WITH UNCOUPLED CONTINUOUS PIECEWISE-LINEAR RESISTOR

    R TOKUNAGA, M KOMURO, T MATSUMOTO, LO CHUA

    INTERNATIONAL JOURNAL OF CIRCUIT THEORY AND APPLICATIONS   17 ( 1 ) 71 - 85  1989.01

    DOI CiNii

  • Bifurcation Analysis of A Cusp-Constrained Piecewise-linear Circuit

    R.Tokunaga, L. O. Chua, T.Matsumoto

    Int. J.CTA   17 ( 3 ) 283 - 346  1989

    DOI CiNii

  • CHAOS + BIFURCATIONS OF CIRCUITS + SYSTEMS - EDITORIAL

    T MATSUMOTO, FMA SALAM

    IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS   35 ( 7 ) 766 - 767  1988.07

    Other  

  • 多重巻き込み:R-L-diode回路にみられるアトラクタの生成メカニズム

    野口淳, 田中聡, 松本隆, L.O.Chua

    電子情報通信学会技術研究報告   CAS88 ( 53 ) 37 - 44  1988

  • Homoclinic Connection II : Double Scroll回路

    堀田篤宏, 藤本竜一, 徳永隆治, 松本隆

    電子情報通信学会技術研究報告   CAS88 ( 58 ) 71 - 77  1988

  • Homoclinic Connections I:区分線形Lorenz回路

    井田孝, 徳永隆治, 松本隆

    電子情報通信学会技術研究報告   CAS88 ( 57 ) 63 - 70  1988

  • Reality of Chaos in the Double Scroll Circuit: A Computer-Assisted Proof

    T.Matsumoto, L. O. Chua, K.Ayaki

    IEEE Trans. CAS   35 ( 7 ) 909 - 925  1988

    DOI CiNii

  • Reality of Chaos in the Double Scroll Circuit: A Computer-Assisted Proof

    T.Matsumoto, L. O. Chua, K.Ayaki

    IEEE Trans. CAS   35 ( 7 ) 909 - 925  1988

    DOI CiNii

  • BIFURCATION SCENARIO IN A DRIVEN R-L-DIODE CIRCUIT

    S TANAKA, T MATSUMOTO, LO CHUA

    PHYSICA D   28 ( 3 ) 317 - 344  1987.10

    DOI CiNii

  • CHAOS IN ELECTRONIC-CIRCUITS

    T MATSUMOTO

    PROCEEDINGS OF THE IEEE   75 ( 8 ) 1033 - 1057  1987.08

    DOI

  • Glendinning-Sparrow Heteroclinic分岐とFishhook構造がDouble Scroll回路で観測された:II

    黒田篤司, 牧瀬哲郎, 徳永隆治, 松本隆

    電子情報通信学会技術研究報告   CAS87 ( 188 ) 41 - 48  1987

  • Glendinning-Sparrow Heteroclinic分岐とFishhook構造がDouble Scroll回路で観測された:I

    黒田篤司, 牧瀬哲郎, 徳永隆治, 松本隆

    電子情報通信学会技術研究報告   CAS87 ( 188 ) 33 - 40  1987

  • 'Lorenz Attractor' from an Electrical Circuit with Uncoupled Piecewise-Linear Resistor

    R.Tokunaga, M.Komuro, T.Matsumoto, L.O.Chua

    Proc. ISCAS   2   672 - 675  1987

  • BIRTH AND DEATH OF THE DOUBLE SCROLL

    T MATSUMOTO, LO CHUA, M KOMURO

    PHYSICA D   24 ( 1-3 ) 97 - 124  1987.01

    DOI CiNii

  • Chaos via Torus Breakdown

    T.Matsumoto, L. O. Chua, R.Tokunaga

    IEEE Trans. CAS   34 ( 3 ) 240 - 253  1987

    DOI CiNii

  • 'Lorenz Attractor' from an Electrical Circuit with Uncoupled Piecewise-Linear Resistor

    R.Tokunaga, M.Komuro, T.Matsumoto, L.O.Chua

    Proc. ISCAS   2   672 - 675  1987

  • HYPERCHAOS - LABORATORY EXPERIMENT AND NUMERICAL CONFIRMATION

    T MATSUMOTO, LO CHUA, K KOBAYASHI

    IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS   33 ( 11 ) 1143 - 1147  1986.11

    Rapid communication, short report, research note, etc. (scientific journal)  

  • THE DOUBLE SCROLL BIFURCATIONS

    T MATSUMOTO, LO CHUA, M KOMURO

    INTERNATIONAL JOURNAL OF CIRCUIT THEORY AND APPLICATIONS   14 ( 2 ) 117 - 146  1986.04

    DOI CiNii

  • 区分線形電子回路における'Lorenz'attractor

    徳永隆治, 小室元政, 松本隆, L.O.Chua

    電子通信学会技術研究報告   CAS86 ( 158 ) 31 - 36  1986

  • 電子回路のchaotic attractor:Double Scroll

    松本隆, 小室元政, L.O.Chua

    月刊フィジックス   7 ( 1 ) 56 - 65  1986

  • Double Scroll回路はchaoticである:厳密証明

    松本隆, 小室元政, L.O.Chua

    電子通信学会技術研究報告   CAS86 ( 159 ) 37 - 42  1986

  • Arnold Tongues

    Tetsuo Makise, Ryuji Tokunaga, Kenichiro Ayaki, Kazuya Tokumasu, Takashi Matsumoto

    Proc. NICOGRAPH'86     116 - 124  1986

  • Poincare Maps of the Double Scroll

    M.Komuro, T.Matsumoto, L. O. Chua

    Dynamical Systems and Nonlinear Oscillation   1   238 - 246  1986

  • The Double Scroll Via A Two-Transistor Circuit

    T.Matsumto, L. O. Chua, K.Tokumasu

    IEEE Trans. CAS   33 ( 8 ) 828 - 835  1986

    DOI CiNii

  • The TF-Equivalence Class Approach to Analog Fault Diagnosis Problems

    Y.Tokawa, T.Matsumoto, H.Arai

    IEEE Trans. CAS   33 ( 10 ) 992 - 1009  1986

    DOI CiNii

  • The Double Scroll Family

    L. O. Chua, M.Komuro, T.Matsumoto

    IEEE Trans. CAS   33 ( 11 ) 1073 - 1118  1986

  • Arnold Tongues

    Tetsuo Makise, Ryuji Tokunaga, Kenichiro Ayaki, Kazuya Tokumasu, Takashi Matsumoto

    Proc. NICOGRAPH'86     116 - 124  1986

  • Poincare Maps of the Double Scroll

    M.Komuro, T.Matsumoto, L. O. Chua

    Dynamical Systems and Nonlinear Oscillation   1   238 - 246  1986

  • The Double Scroll Via A Two-Transistor Circuit

    T.Matsumto, L. O. Chua, K.Tokumasu

    IEEE Trans. CAS   33 ( 8 ) 828 - 835  1986

    DOI CiNii

  • The TF-Equivalence Class Approach to Analog Fault Diagnosis Problems

    Y.Tokawa, T.Matsumoto, H.Arai

    IEEE Trans. CAS   33 ( 10 ) 992 - 1009  1986

    DOI CiNii

  • The Double Scroll Family

    L. O. Chua, M.Komuro, T.Matsumoto

    IEEE Trans. CAS   33 ( 11 ) 1073 - 1118  1986

  • The Double Scroll

    T.Matsumoto, L.O.Chua, M.Komuro

    IEEE TRANS. CAS   32 ( 8 ) 797 - 818  1985

    DOI CiNii

  • Bifurcations of a Driven R-L-Diode Circuit

    S.Tanaka, T.Matsumoto, L. O. Chua

    Proc. IEEE ISCAS     851 - 854  1985

  • 3次元autonomous回路のtorus崩壊

    徳永隆治, 松本隆

    電子通信学会技術研究報告   CAS-85 ( 123 ) 25 - 32  1985

  • Bifurcations of the Double-Scroll

    T.Matsumoto, L.O.Chua, M.Komuro

    Proc. IEEE ISCAS   1   175 - 178  1985

  • Double ScrollのTakens再構成: 実験と検証

    黒川誠, 小川和人, 徳升一也, 牧瀬哲郎, 松本隆

    電子通信学会技術研究報告   CAS-85 ( 122 ) 17 - 24  1985

  • Bifurcations of the Double Scroll

    T. Matsumoto

    Proceedings of the 24th IEEE Conference on Decision and Control   1   455  1985

  • The Double Scroll

    T.Matsumoto, L.O.Chua, M.Komuro

    IEEE TRANS. CAS   32 ( 8 ) 797 - 818  1985

    DOI CiNii

  • Bifurcations of a Driven R-L-Diode Circuit

    S.Tanaka, T.Matsumoto, L. O. Chua

    Proc. IEEE ISCAS     851 - 854  1985

  • Bifurcations of the Double-Scroll

    T.Matsumoto, L.O.Chua, M.Komuro

    Proc. IEEE ISCAS   1   175 - 178  1985

  • Bifurcations of the Double Scroll

    T. Matsumoto

    Proceedings of the 24th IEEE Conference on Decision and Control   1   455  1985

  • 2-Segment Piecewode-linearキャパシタ回路の非周期的アトラクタ

    田中 聡, 松本 隆, L. O. Chua

    電子通信学会技術研究報告   CAS84-168   47 - 51  1984

  • 3階相反回路に見られる非周期解:Double-Scroll アトラクタ

    松本隆, 小室元政, L.O.Chua

    電子通信学会技術研究報告   CAS84-169   53 - 60  1984

  • A Chaotic Attractor from an Autonomous 3-Segment Piecewise-Linear Circuit

    T.Matsumoto, L. O. Chua, M.Komuro

    Theory of Dynamical Systems and Its Applications to Nonlinear Problems     194 - 198  1984

  • A Chaotic Attractor from a 2-Segment Piecewise- Linear Capacitor Circuit

    S.Tanakam, T.Matsumoto, L. O. Chua

    Theory of Dynamical Systems and Its Applications to Nonlinear Problems     181 - 193  1984

  • On the Topological Testability Conjecture for Analog Fault Diagnosis Problems

    Y.Togawa, T.Matsumoto

    IEEE Trans. CAS   31 ( 2 ) 147 - 158  1984

    DOI CiNii

  • Simplest Chaotic Non-Autonomous Circuit

    T.Matsumoto, L. O. Chua, S.Tanaka

    Phys. Rev. A   30 ( 2 ) 1155 - 1157  1984

    DOI CiNii

  • A Chaotic Attractor from Chua's Circuit

    T.Matsumoto

    IEEE Trans,Circuit & Systems   31 ( 12 ) 1055 - 1058  1984

    DOI CiNii

  • A Chaotic Attractor from an Autonomous 3-Segment Piecewise-Linear Circuit

    T.Matsumoto, L. O. Chua, M.Komuro

    Theory of Dynamical Systems and Its Applications to Nonlinear Problems     194 - 198  1984

  • A Chaotic Attractor from a 2-Segment Piecewise- Linear Capacitor Circuit

    S.Tanakam, T.Matsumoto, L. O. Chua

    Theory of Dynamical Systems and Its Applications to Nonlinear Problems     181 - 193  1984

  • On the Topological Testability Conjecture for Analog Fault Diagnosis Problems

    Y.Togawa, T.Matsumoto

    IEEE Trans. CAS   31 ( 2 ) 147 - 158  1984

    DOI CiNii

  • Simplest Chaotic Non-Autonomous Circuit

    T.Matsumoto, L. O. Chua, S.Tanaka

    Phys. Rev. A   30 ( 2 ) 1155 - 1157  1984

    DOI CiNii

  • A Chaotic Attractor from Chua's Circuit

    T.Matsumoto

    IEEE Trans,Circuit & Systems   31 ( 12 ) 1055 - 1058  1984

    DOI CiNii

  • Strong Structural Stability of Resistive Nonlinear n-Ports

    T. Matsumoto, G. Ikegami, L. O. Chua

    IEEE Trans., CAS   30 ( 4 ) 197 - 222  1983

    DOI CiNii

  • Strong Structural Stability of Resistive Nonlinear n-Ports

    T. Matsumoto, G. Ikegami, L. O. Chua

    IEEE Trans., CAS   30 ( 4 ) 197 - 222  1983

    DOI CiNii

  • 1982 ISCAS報告(IV)非線形回路-Nonlinear Networks

    松本隆

    電子通信学会技術研究報告   CAS82 ( 121 ) 77 - 82  1982

  • 非線形抵抗n-Port 強構造安定性について

    松本 隆, 池上宣弘, L. O. Chua

    電子通信学会技術研究報告   CAS-82-138 ( 208 ) 93 - 98  1982

  • Strong Structural Stability of Resistive Nonlinear n-ports

    T. Matsumoto, G. Ikegami, L. O. Chua

    IEEE International Symposium on Circuits and Systems   2   455 - 458  1981

  • GEOMETRIC-PROPERTIES OF DYNAMIC NON-LINEAR NETWORKS - TRANSVERSALITY, LOCAL-SOLVABILITY AND EVENTUAL PASSIVITY

    T MATSUMOTO, LO CHUA, H KAWAKAMI, S ICHIRAKU

    IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS   28 ( 5 ) 406 - 428  1981

    DOI CiNii

  • 非線形解析の動向

    松本隆

    電子通信学会論文誌   63 ( 5 ) 511 - 513  1980

  • GEOMETRIC-PROPERTIES OF RESISTIVE NON-LINEAR N-PORTS - TRANSVERSALITY, STRUCTURAL STABILITY, RECIPROCITY, AND ANTI-RECIPROCITY

    LO CHUA, T MATSUMOTO, S ICHIRAKU

    IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS   27 ( 7 ) 577 - 603  1980

    DOI CiNii

  • 電気回路網の安定性について

    松本隆, 一楽重雄, 川上博, L.O.Chua

    京大数理解析研究所講究録,力学系における非線形回路の諸問題   370   30 - 43  1979

  • On the Implications of Capacitor-only Cutsets and Inductor-only Loops

    T.Matsumoto, L. O. Chua, M.Makino

    IEEE Trans CAS   26 ( 10 ) 828 - 845  1979

    DOI CiNii

  • On the Implications of Capacitor-only Cutsets and Inductor-only Loops

    T.Matsumoto, L. O. Chua, M.Makino

    IEEE Trans CAS   26 ( 10 ) 828 - 845  1979

    DOI CiNii

  • An Implicit Function Theorem and Its Applications to Nonlinear Electrical Networks

    T.Matsumoto

    SIAMJ. Math. Anal   9 ( 3 ) 492 - 499  1978

  • 周期解を持つ非線形RC相反回路網について

    松本隆, 川上博

    電子通信学会論文誌   J61-A ( 1 ) 71 - 72  1978

  • 非線形回路網の実質強受動性とコンパクトアトラクション集合について

    松本隆, 一楽重雄

    電子通信学会論文誌   J61-A ( 10 ) 925 - 932  1978

  • An Implicit Function Theorem and Its Applications to Nonlinear Electrical Networks

    T.Matsumoto

    SIAMJ. Math. Anal   9 ( 3 ) 492 - 499  1978

  • Passivity and Eventual Passivity of Electrical Networks

    T.Matsumoto

    Academic Press     459 - 463  1977

  • Eventual Passivity of Non-Linear Resistive Networks and their Operating Points”

    T. Matsumoto, T. Sato

    Electronics & Communications in Japan 1977   60 ( 2 ) 51 - 59  1977

  • Eventual Passivity of Nonlinear Resistive Networks and Their Operating Points

    T. Matsumoto, T. Sato

    Trans. of the Institute of Electronics and Communication Engineers of Japan   E60 ( 2 ) 88 - 89  1977

  • 非線形抵抗回路網の実質受動性とその動作点について

    松本隆, 佐藤俊明

    電子通信学会論文誌   60-A ( 2 ) 184 - 191  1977

  • Eventually Passive Nonlinear Networks

    T.Matsumoto

    IEEE Trans. CAS   24 ( 5 ) 261 - 269  1977

    DOI CiNii

  • Twin-Tを含むある種の能動RCフィルタにおける縮退維持について

    松本隆, 山本恒夫

    電子通信学会論文誌   J60-A ( 6 ) 587 - 588  1977

  • 実質強受動性について

    松本隆, 一楽重雄

    京都大学数理解析研究所講究録   313   59 - 69  1977

  • Passivity and Eventual Passivity of Electrical Networks

    T.Matsumoto

    Academic Press     459 - 463  1977

  • Eventual Passivity of Non-Linear Resistive Networks and their Operating Points”

    T. Matsumoto, T. Sato

    Electronics & Communications in Japan 1977   60 ( 2 ) 51 - 59  1977

  • Eventual Passivity of Nonlinear Resistive Networks and Their Operating Points

    T. Matsumoto, T. Sato

    Trans. of the Institute of Electronics and Communication Engineers of Japan   E60 ( 2 ) 88 - 89  1977

  • Eventually Passive Nonlinear Networks

    T.Matsumoto

    IEEE Trans. CAS   24 ( 5 ) 261 - 269  1977

    DOI CiNii

  • 電気回路網におけるエネルギー、パワーそして混合ポテンシャルについて

    松本隆

    京大数理解析研究所講究所講究録,電気回路の力学系   284   1 - 17  1976

  • Some Properties of Min-Max Functions

    T.Matsumoto

    SIAM Journal on Control 1976   14 ( 1 ) 144 - 155  1976

    DOI

  • Eventual Passivity of Nonlinear Networks

    T. Matsumoto, T. Sato

    Electronics Communications in Japan 1976   59 ( 9 ) 24 - 32  1976

  • Dynamical Systems Arising from Electrical Networks

    T.Matsumoto

    Dynamical Systems     285 - 290  1976

  • On the Dynamics of Electrical Networks

    T.Matsumoto

    J. Differential Equations   21 ( 1 ) 179 - 196  1976

  • On Several Geometric Aspects of Nonlinear Networks

    T.Matsumoto

    J.Franklin Institute   301 ( 1,2 ) 203 - 225  1976

  • On a Class of Nonlinear Networks

    T.Matsumoto

    Int. J. Cir. Theor. Appl.   4 ( 1 ) 55 - 73  1976

    DOI CiNii

  • On Some Properties of Min-Max Functions

    T.Matsumoto

    SIAM J. Control and Optimization   14 ( 1 ) 144 - 155  1976

    DOI CiNii

  • On Eventual Passivity of Nonlinear Networks

    T. Matsumoto, T. Sato

    Trans. of the Institute of Electronics & Communication Engineers of Japan   E59 ( 9 ) 22 - 23  1976

  • 非線形回路網の実質受動性について

    松本隆, 佐藤俊明

    電子通信学会論文誌   59-A ( 9 ) 718 - 725  1976

  • Some Properties of Min-Max Functions

    T.Matsumoto

    SIAM Journal on Control 1976   14 ( 1 ) 144 - 155  1976

    DOI

  • Eventual Passivity of Nonlinear Networks

    T. Matsumoto, T. Sato

    Electronics Communications in Japan 1976   59 ( 9 ) 24 - 32  1976

  • Dynamical Systems Arising from Electrical Networks

    T.Matsumoto

    Dynamical Systems     285 - 290  1976

  • On the Dynamics of Electrical Networks

    T.Matsumoto

    J. Differential Equations   21 ( 1 ) 179 - 196  1976

  • On Several Geometric Aspects of Nonlinear Networks

    T.Matsumoto

    J.Franklin Institute   301 ( 1,2 ) 203 - 225  1976

  • On a Class of Nonlinear Networks

    T.Matsumoto

    Int. J. Cir. Theor. Appl.   4 ( 1 ) 55 - 73  1976

    DOI CiNii

  • On Some Properties of Min-Max Functions

    T.Matsumoto

    SIAM J. Control and Optimization   14 ( 1 ) 144 - 155  1976

    DOI CiNii

  • On Eventual Passivity of Nonlinear Networks

    T. Matsumoto, T. Sato

    Trans. of the Institute of Electronics & Communication Engineers of Japan   E59 ( 9 ) 22 - 23  1976

  • Class of Nonlinear Networks

    T.Matsumoto

    Electronics & Communications in Japan 1975   58 ( 7 ) 59 - 67  1975

  • Several Qualitative Aspects of Dynamics of Nonlinear Networks

    T. Matsumoto

    Electronics & Communications in Japan   58 ( 10 ) 13 - 21  1975

  • 電気回路網のダイナミクスについて

    松本隆

    京大数理解析研究所講究録 電気回路の力学系   254   63 - 88  1975

  • On Some Sensitivity Formulas for Networks in Frequency-Domain

    T.Matsumoto

    Electronics & Communications in Japan 1975   58 ( 2 ) 48 - 56  1975

  • 線形逃避問題

    松本隆

    計測自動制御学会論文集   11 ( 1 ) 63 - 69  1975

    DOI

  • 周波数領域における回路網のいくつかの素子感度公式について

    松本隆

    電子通信学会論文誌   58-A ( 2 ) 129 - 136  1975

  • 大規模系のいくつかの性質

    松本隆

    計測自動制御学会論文集   11 ( 2 ) 125 - 131  1975

    DOI

  • 回路網の最適化における随伴方程式について

    松本隆

    電子通信学会論文誌   58-A ( 4 ) 234 - 236  1975

  • あるクラスのmin-max関数

    松本隆

    計測自動制御学会論文集   11 ( 3 ) 322 - 327  1975

    DOI

  • 非線形回路網のエネルギー関数について

    松本隆

    電子通信学会論文誌   58-A ( 7 ) 467 - 468  1975

  • A Class of Linear Evasion Problems

    T.Matsumoto

    J. Opt. Theor. & Appl   16 ( 1 ) 147 - 164  1975

  • 非線形回路網の幾つかの性質について

    松本隆

    電子通信学会論誌   58-A ( 10 ) 617 - 624  1975

  • 非線形回路網のダイナミクスの定性的考察

    松本隆

    電子通信学会論文誌   58-A ( 10 ) 625 - 632  1975

  • アクティブフィルタの極指定の一手法

    松本隆

    電子通信学会論文誌   58-A ( 11 ) 720 - 721  1975

  • Class of Nonlinear Networks

    T.Matsumoto

    Electronics & Communications in Japan 1975   58 ( 7 ) 59 - 67  1975

  • Several Qualitative Aspects of Dynamics of Nonlinear Networks

    T. Matsumoto

    Electronics & Communications in Japan   58 ( 10 ) 13 - 21  1975

  • On Some Sensitivity Formulas for Networks in Frequency-Domain

    T.Matsumoto

    Electronics & Communications in Japan 1975   58 ( 2 ) 48 - 56  1975

  • A Class of Linear Evasion Problems

    T.Matsumoto

    J. Opt. Theor. & Appl   16 ( 1 ) 147 - 164  1975

  • Generalization of Implicit Function Theorem and it’s applications to nonlinear Networks

    T.Matsumoto

    Electronics & Communications in Japan 1974   57 ( 10 ) 40 - 48  1974

  • Gradients of a Performance Index Arising from Network Optimization in the Frequency Domain

    T.Matsumoto

    Electronics Letters   10 ( 13 ) 263  1974

  • あるクラスの非線形回路網について

    松本隆

    電子通信学会論文誌   57-A ( 7 ) 535 - 542  1974

  • 陰関数定理の一つの一般化とその非線形回路網への応用について

    松本隆

    電子通信学会論文誌   57-A ( 10 ) 738 - 745  1974

  • 極値をとる関数の性質及びそのペイズ決定問題への応用

    松本隆

    計測自動制御学会論文集   10 ( 6 ) 657 - 661  1974

    DOI

  • Generalization of Implicit Function Theorem and it’s applications to nonlinear Networks

    T.Matsumoto

    Electronics & Communications in Japan 1974   57 ( 10 ) 40 - 48  1974

  • Gradients of a Performance Index Arising from Network Optimization in the Frequency Domain

    T.Matsumoto

    Electronics Letters   10 ( 13 ) 263  1974

  • On a Class of Linear Pursuit Games

    T.Matsumoto, E.Shimemura

    J. of Differential Equations   11 ( 2 ) 266 - 283  1972

  • Online SSVEP-Based Brain-Machine Interface With Automatic Determination of Stopping Time of Training Phase

    Y.Dobashi, A.Takemoto, S.Shigezumi, T. Shiraki, K.Nakamura, T.Matsumoto

    International Journal of Computational Bioscience   accepted

  • Online SSVEP-Based Brain-Machine Interface With Automatic Determination of Stopping Time of Training Phase

    Y.Dobashi, A.Takemoto, S.Shigezumi, T. Shiraki, K.Nakamura, T.Matsumoto

    International Journal of Computational Bioscience   accepted

▼display all

Industrial Property Rights

  • オンライン手書き文字認識方法

    Patent

 

Overseas Activities

  • 階層Bayes的学習・予測手法の新展開とImplementation

    2003.09
    -
    2004.09

    イギリス   ケンブリッジ大学

Internal Special Research Projects

  • NIRS/EEG/HIマルチモーダル・アメニティ指標の構築

    2009  

     View Summary

    【全体像】表1はEEG, NIRS, そしてHI(hyperspectral imaging)の得失を列挙したものである。これら)を組み合わせたマルチモーダル・アメニティ指標を構築に挑戦を試みることを目的とした。           EEG    NIRS   HI  時間分解能  高い   低い   低い  空間分解能  低い   高い   高い  装置規模   中・小  中・小  中・大            表1 【EEG】まずEEGについてはSSVEPによる注視と、selected attentionによる刺激周波数成分検出を行い、前者で98%、後者で70%の判別率を得た。正準変数を抽出し、非線形判別器を用い、Bayes的枠組みからMCMC(Markov Chain Monte Carlo)で実装した。図1、図2に周波数成分と典型的な判別データを示す。【HI】 HIを用い、主として手と腕の画像を波長200nm-1000nmまで約200枚を採取し、現在精査中である。【NIRS】NIRSでは、4種類の匂い(ethanol, geranium,jasmine,peppermint)をかいだときの前頭前野の酸化ヘモグロビン分布をNIRSで計測した。図3は、jasumineをかいだときの酸化ヘモグロビン量をpsuedo colar表示したものである。興奮性の匂いであることが見て取れる。どの匂いについても約01Hzの基本振動が観測されたが、別途行ったEEGではそのような周波数の振動は観測されなかった。この研究内で検討すべきか否かは別として、きわめて興味深い現象と考えている。 時間・予算の拘束が厳しく、これらを統合するにはいたらなかったが、次年度以後、挑戦していきたいと考えている。

  • NIRSによる脳情報解読オンラインアルゴリズムの構築

    2009  

     View Summary

    【ターゲット】4種類の匂い(ethanol, geranium, jasmine, peppermint)をかいだときの「前頭前野」の「酸化ヘモグロビン」分布をNIRSで計測した。【実験】被験者にゆったり座ってもらい、全体が安定したと思われる状況を確認後、まず5秒間当該の匂いをかいでもらった後、30秒のレスト期間をもうけ、その後再び5秒間の匂いかぎ、そして30秒のレスト期間の後、もう一度匂いかぎとレスト期間、という手順を踏み、次のような知見を得た: 1.どの匂いの場合も、約0.1Hzの基本振動を含んでいると思われる。 2.一方、別の実験によるEEGデータではこのような0.1Hz振動は観測されなかった。 3.Ethanolの場合、酸化ヘモグロビンは減少の傾向があり、これは神経細胞が脱分極していることを意味しているので、「沈静作用」をもつと思われる。 4.Jasmineの場合、逆に酸化ヘモグロビンは増加傾向にあり、「興奮作用」を及ぼすと思われる。図1は典型的なjasumine観測データである。 5.後の二つは今回の実験に関する限り、明快な結論には到達しなかった。【オンラインアルゴリズム】 Bayes的枠組みからオンライン学習アルゴリズムを構築、それを「逐次モンテカルロ」で実装する方策を鋭意進めている。図1jasmineを嗅いだときの典型的NIRS応答赤:酸化ヘモグロビン、青:脱酸化ヘモグロビン、青:全体匂いかぎ開始:t=100、終了:t=150

  • 逐次周辺尤度オンライン変化検出:粒子フィルタ的接近

    2005  

     View Summary

    工学における「変化検出」(change detection) は、興味の対象としている広い意味のシステム内部の突発的変化の検出を意味するのが一般的である。“突発的変化”は瞬時的あるいはデータ収集速度より遥かに速い変化を意味する。従ってこれはいわゆる適応推定 (adaptive estimation) とは異なる範疇の問題である。適応推定ではシステムのパラメータが緩やかに変化する場合である。変化が突発的であることは変化量が大きい事を意味しない。むしろ逆に変化量が小さい場合がチャレンジングである。工学で扱う問題ではデータに不確定性 (雑音である事が多い) が含まれているのが普通なのでそのような枠組みで問題を捉える必要がある。 優れた変化検出アルゴリズムは広大な応用を持つ。広い意味のシステムにおける故障検出、不正検出、動画像における場面変化検出、話者変化検出、医用データ変化検出、製品品質変化検出、環境データ変化検出、天候変化検出、地理データ変化検出、等々である。 変化検出アルゴリズムは大きく分けて二つのクラスに分けることができる。ひとつはモデル準拠型、そしてもうひとつはデータ準拠型である。前者では対象としているシステムを既述する方程式が分かっている場合、後者はそれが分からない、あるいは原理としては方程式導出が可能なはずであっても現実には極めて困難な場合であって、得られたたデータのみから変化を検出しようとする手法である。この研究はデータ準拠手法の一つであり、Bayes的オンライン学習の枠組みから興味のある変数、例えば「逐次周辺尤度」、隠れ変数としての「変化変数」など、の逐次事後分布公式を導き、実装を「逐次モンテカルロ(Sequential Monte Carlo)」手法(粒子フィルタ、Particle Filterとも呼ばれる)で遂行している。すでに幾つかの成果を生んでおり、国際雑誌、国際会議、などで採録されている。

  • Pecora-Carrol カオス同期とその応用可能性の検討に関する研究

    1997  

     View Summary

    1. R-L-Diode 回路はその単純さにもかかわらず、豊かな分岐現象( 周期倍分岐、saddle-node 分岐、intermittency, crisis, chaos 等 ) が観測される。我々の様な工学の分野だけでなく物理学の研究者達もこの回路を研究している理由の一つはこの系には人工的 ( artificial) な部分が全くない自然な系だからと思われる。 Pecora-Carrol カオス同期と関連してchaotic masking というパラダイムが提案されている [Cosmo and Oppenheim, Phys. Rev. Lett. 1993]。これはカオス的ふるまいを示す系の状態変数で、情報信号(例えば音声)をおおってしまい(masking)それを通信系を通して送り、受信側で適当な同期系を構成してもとの情報信号を復元しようとするものである。本研究で検討しているR-L-Diode 回路で情報信号を音楽として実験を行い、復号可能性を昨年度確認したが、それに附随して新しいopen problemが二つ発生した:(i)“復号”された信号はもとの情報信号(音楽)の振巾の約70倍となり、カオス同期系に増巾メカニズムが組み込まれている事がわかった。が、何故増巾が起こるかは未解決である。(ii) masking schemeが何故働くかも未解決である。2. 第一年度に得たカオス同期の理論的正当性は、R-L-Diode回路に特化したものであったが、より広いクラスの一般的非線形ダイナミカルシステムへも適用可能である事を示した。研究成果の発表[1] 「電子回路に見るカオス、分岐、同期」, 電子情報通信学会、電子デバイス研究会、電子情報通信学会技術報告、 ED-96-216, pp. 1 - 8,March 14, 1997,著者:松本 隆、西 正信[2] 「RーLーDiode 回路におけるPecora-Carrol 同期回路の単純化」、1997年電子情報通信学会総合大会、A-2-46, pp.103, March, 1997,著者:浜野英知、西 正信、松本 隆[3] “R-L-Diode 回路におけるTopological Horseshoe: A Computer Assisted Proof Topological Horseshe in the R-L-Diode circuit”, 電気学会研究会、情報処理研究会、IP-97-13, p.1 - 10, Sept. 1997,著者:竹内規晃、松本 隆[4] 「R-L-Diode回路におけるTopological Horseshoe: Computer Assisted Proof」, 1997年電子情報通信学会基礎・境界ソサイエティ大会,A-2-1, p.29, Sept. 1997,著者:竹内規晃、松本 隆[5] 「R-L-Diode回路のPecora-Carroll カオス同期とマスキング」、電子情報通信学会論文誌A, vol. J80-A, No. 9, pp. 1421-1430, Sept. 1997,著者:西 正信、松本 隆[6] 「R-L-Diode回路のPecora-Carroll カオス同期とマスキング」、1997年電子情報通信学会基礎・境界ソサイエティ大会、A-2-2,p.30, Sept. 1997,著者:西 正信、松本 隆[7] “Chaos, Synchronization and Bifurcations in a Driven R-L-Diode Circuit”, Proc. IUTAM Chaos '97, Kluwer-Acedemic 著者:西 正信、松本 隆[8] 「非線形ダイナミカルシステムの部分システム単調性とカオス同期」、電子情報通信学会技術研究報告、vol.97, No.530, NLP-139, pp.25-32, Feb.1998,著者:西 正信、松本 隆 [9] 「非線形ダイナミカルシステムの部分システム単調性とカオス同期」、電子情報通信学会誌A, to appear著者:西 正信、浜野英知、松本 隆[10] 「R-L-Diode回路におけるTopological Horseshoeについて」、電子情報通信学会誌A, to appear著者:竹内規晃、永井立紀、松本 隆[11] “Subsystem Decreasing for Exponential Synchronization of Chaotic Systems”, Physical Review E, to appear,著者:T. Matsumoto, M. Nishi

  • R-L-Diode回路のカオス同期:Secure通信への応用可能性

    1996  

     View Summary

     1990年、米国の物理学者PecoraとCarroll{Pecora and Carroll,Phys.Rev.Lett.64,p821}はchaoticな系を部分系に分割し、分割された部分系が、もとのchaoticな系と同期(synchronize)する条件を調べた。これをもとに、1993年米国の電気工学者Cuomo Oppenheim{Cuomo and Oppenheim,Phys.Rev.Lett,71,p.65}は、Lorenz方程式をアナログ回路で組み、chaoticな状態を作り出しておき、それにアナログ又はディジタル信号をencodeして信号を送り、受信側でchaotic synchronizationによってdecodeする実験を行い注目されている。伝送される信号はchaoticなので、たとえ盗聴者があっても、対応するLorenz方程式を実現する回路とそのパラメータを知らなければdecodeする事な不可能である。勿論、膨大な計算時間を費やせば、系の同定は不可能ではないが、比較的短時間のsecure通信のためには十分安全であると言えよう。このアイデアは未だ全く初期的段階であるが、考え方としては非常に面白いものが含まれている。 本研究の目的は本申請者が長年調べてきた単純なchaotic回路(RL Diode回路)を用いてchaotic synchronizationとそのsecure通信への応用の可能性を探る事である。 この回路はLorenz方程式を実現する回路に比べて圧倒的に単純であり本格的implemetationの際決定的なfactorとなり得よう。また、Lorenz回路は所謂自律系であるのに対して、R-L-Diode回路は非自律系である事もLorentz系の場合とは異なる側面の一つである。前年度までに次の結果を得ている: 1.Pecora-Carrolカオス同期の実験的観測。 2.理論的結果による実験の検証。 3.Chaotic Maskingの実験により音楽波形の復元。今年度は次の課題を解決した。 4.受信側で独立電圧源に関する位相情報がない場合についてもPLLとフィルタを用いてPecora-Carrol同期を実験的に確認した。 5.対象としているR-L-Diode回路の大域分岐を解明し、カオス同期現象を調べる基礎を与えた。

  • R-L-Diode回路のカオス同期とそのSecure通信への応用可能性

    1995  

     View Summary

    1990年,米国の物理学者PecoraとCarroll〔Pecora and Carroll, Phys. .Rev. Lett., 64, p. 821〕はchaoticな系を部分的に分割し,分割された部分系が,もとのchaoticな系と同期(synchronize)する条件を調べた。これをもとに,1993年米国の電気工学者CuomoとOppenheim〔Cuomo and Oppenheim, Phys. Rev. Lett., 71, p.65〕は,Lorenz方程式をアナログ回路で組み,chaoticな状態を作り出しておき,それにアナログ又はディジタル信号をencodeして信号を送り,受信側でchaotic synchronizationによってdecodeする実験を行い注目されている。伝送される信号はchaoticなので,たとえ盗聴者があっても,対応するLorenz方程式を実現する回路とそのパラメータを知らなければdecodeする事は不可能である。勿論,膨大な計算時間を費やせば,系の同定は不可能ではないが,比較的短時間のsecure通信のためには十分安全であるといえよう。このアイデアは未だ全く初期的段階であるが,考え方としては非常におもしろいものが含まれている。 本研究の目的は本申請者が長年調べてきた単純なchaotic回路(RL Diode回路)を用いてchaoticsynchronizationとそのsecure通信への応用の可能性を探る事である。 この回路はLorenz方程式を実現する回路に比べて圧倒的に単純であり,本格的implementationの際決定的なfactorとなり得よう。またLorenz回路は所謂自律系であるのに対して,R-L-Diode回路は非自律系である事もLorenz系の場合とは異なる側面の一つである。 所期の目的は完全に達成された。具体的には1. Pecoa-Carrolカオス同期の実験的観測をおこなった。2. 理論的結果を与え,実験を検証した。3. Chaotic Maskingの実験を行い,音楽波形の復元に成功した。

▼display all