KATTO, Jiro

写真a

Affiliation

Faculty of Science and Engineering, School of Fundamental Science and Engineering

Job title

Professor

Homepage URL

http://www.katto.comm.waseda.ac.jp/

Concurrent Post 【 display / non-display

  • Faculty of Science and Engineering   Graduate School of Fundamental Science and Engineering

  • Affiliated organization   Global Education Center

Research Institute 【 display / non-display

  • 2020
    -
    2022

    国際情報通信研究センター   兼任研究員

  • 2020
    -
    2022

    理工学術院総合研究所   兼任研究員

Degree 【 display / non-display

  • Ph.D

Research Experience 【 display / non-display

  • 2008
    -
     

    Waseda University   School of Fundamental Science and Engineering

  • 2004
    -
    2008

    Professor, Waseda University

  • 2004
    -
    2008

    Director, NEDO

  • 1999
    -
    2004

    Associate Professor, Waseda University

  • 1992
    -
    1999

    Researcher, NEC C&C Laboratories

display all >>

Professional Memberships 【 display / non-display

  •  
     
     

    ACM

  •  
     
     

    IEEE

  •  
     
     

    画像電子学会

  •  
     
     

    映像情報メディア学会

  •  
     
     

    情報処理学会

display all >>

 

Research Areas 【 display / non-display

  • Communication and network engineering

  • Computer system

Research Interests 【 display / non-display

  • Multimedia Signal Processing

  • Computer Networks

Papers 【 display / non-display

  • IoT-centric service function chainingorchestration and its performance validation

    Hibiki Sekine, Kenji Kanai, Jiro Katto, Hidehiro Kanemitsu, Hidenori Nakazato

    2021 IEEE 18th Annual Consumer Communications and Networking Conference, CCNC 2021    2021.01

     View Summary

    In order to simplify deployment and management of IoT services, Network Function Virtualization (NFV) and Service Function Chaining (SFC) are promising solutions, and much researchers have conducted these topics. To enhance the reliability of former research efforts, in this paper, we propose an orchestration framework for IoT-centric SFC by using Docker and Kubernetes. The framework enables an automatic IoT service deployment by satisfying service requirements and computing and network resource constraints. In such deployment, we apply a Virtual Network Function (VNF)/Service Function (SF) placement problem to achieve efficient utilization of the resources. We set an objective function as minimizing both numbers of SF instances and communications and build a mathematical model based on Integer Linear Programming (ILP). To validate it, we implement a model for the framework and evaluate the performances by carrying out a numerical evaluation and a real experiment. From the evaluation results, we confirm that the proposed approach can reduce the number of SF placements and the number of communications among SF instances.

    DOI

  • Performance evaluations of channel estimation using deep-learning based super-resolution

    Daiki Maruyama, Kenji Kanai, Jiro Katto

    2021 IEEE 18th Annual Consumer Communications and Networking Conference, CCNC 2021    2021.01

     View Summary

    Thanks to breakthrough and evolution of deep learning in computer vision areas, adaptation of deep learning into communication systems are getting lots of attention to researchers. Recently, a channel estimation method by using a deep learning-based image super-resolution (SR) technique, namely ChannelNet, has been proposed. Inspired by this research, in this paper, we propose a deep SR based channel estimation method by applying more accurate deep learning-based SR network architecture, EDSR. In order to enhance intelligibility and reliability of deep SR based channel estimation methods, we evaluate the performance of several deep SR based channel estimation methods (SRCNN, ChannelNet and EDSR) by carrying out practical 5G simulations. From the evaluations, the results conclude that the deep SR based channel estimation methods can potentially improve accuracy of channel estimation and reduce BER characteristics.

    DOI

  • Fully Neural Network Mode Based Intra Prediction of Variable Block Size

    Heming Sun, Lu Yu, Jiro Katto

    2020 IEEE International Conference on Visual Communications and Image Processing, VCIP 2020     21 - 24  2020.12

     View Summary

    Intra prediction is an essential component in the image coding. This paper gives an intra prediction framework completely based on neural network modes (NM). Each NM can be regarded as a regression from the neighboring reference blocks to the current coding block. (1) For variable block size, we utilize different network structures. For small blocks 4×4 and 8×8, fully connected networks are used, while for large blocks 16×16 and 32×32, convolutional neural networks are exploited. (2) For each prediction mode, we develop a specific pre-trained network to boost the regression accuracy. When integrating into HEVC test model, we can save 3.55%, 3.03% and 3.27% BD-rate for Y, U, V components compared with the anchor. As far as we know, this is the first work to explore a fully NM based framework for intra prediction, and we reach a better coding gain with a lower complexity compared with the previous work.

    DOI

  • Enhanced Intra Prediction for Video Coding by Using Multiple Neural Networks

    Heming Sun, Zhengxue Cheng, Masaru Takeuchi, Jiro Katto

    IEEE Transactions on Multimedia   22 ( 11 ) 2764 - 2779  2020.11

     View Summary

    This paper enhances the intra prediction by using multiple neural network modes (NM). Each NM serves as an end-To-end mapping from the neighboring reference blocks to the current coding block. For the provided NMs, we present two schemes (appending and substitution) to integrate the NMs with the traditional modes (TM) defined in high efficiency video coding (HEVC). For the appending scheme, each NM is corresponding to a certain range of TMs. The categorization of TMs is based on the expected prediction errors. After determining the relevant TMs for each NM, we present a probability-Aware mode signaling scheme. The NMs with higher probabilities to be the best mode are signaled with fewer bits. For the substitution scheme, we propose to replace the highest and lowest probable TMs. New most probable mode (MPM) generation method is also employed when substituting the lowest probable TMs. Experimental results demonstrate that using multiple NMs will improve the coding efficiency apparently compared with the single NM. Specifically, proposed appending scheme with seven NMs can save 2.6%, 3.8%, and 3.1% BD-rate for Y, U, and V components compared with using single NM in the state-of-The-Art works.

    DOI

  • HEVC video coding with deep learning based frame interpolation

    Joi Shimizu, Zhengxue Cheng, Heming Sun, Masaru Takeuchi, Jiro Katto

    2020 IEEE 9th Global Conference on Consumer Electronics, GCCE 2020     433 - 434  2020.10

     View Summary

    Recent researches in video frame interpolation show great progress. In this paper, we propose a novel video compression method which incorporates deep learning based frame interpolation into HEVC which is the current video compression standard. Experimental results show that our approach can outperform HEVC in some scenarios.

    DOI

display all >>

Books and Other Publications 【 display / non-display

  • IT Text: インターネットプロトコル

    分担

    オーム社  2005.10

  • ディジタル放送教科書

    分担

    IDGジャパン  2003.02

  • H.323/MPEG-4教科書

    分担

    IEインスティチュート  2001.04

Misc 【 display / non-display

  • Perceptual Quality Study on Deep Learning Based Image Compression

    Zhengxue Cheng, Pinar Akyazi, Heming Sun, Jiro Katto, Touradj Ebrahimi

    Proceedings - International Conference on Image Processing, ICIP   2019-September   719 - 723  2019.09

     View Summary

    Recently deep learning based image compression has made rapid advances with promising results based on objective quality metrics. However, a rigorous subjective quality evaluation on such compression schemes have rarely been reported. This paper aims at perceptual quality studies on learned compression. First, we build a general learned compression approach, and optimize the model. In total six compression algorithms are considered for this study. Then, we perform subjective quality tests in a controlled environment using high-resolution images. Results demonstrate learned compression optimized by MS-SSIM yields competitive results that approach the efficiency of state-of-the-art compression. The results obtained can provide a useful benchmark for future developments in learned image compression.

    DOI

  • Deep Residual Learning for Image Compression

    Zhengxue Cheng, Heming Sun, Masaru Takeuchi, Jiro Katto

       2019.06

    Internal/External technical report, pre-print, etc.  

     View Summary

    In this paper, we provide a detailed description on our approach designed for<br />
    CVPR 2019 Workshop and Challenge on Learned Image Compression (CLIC). Our<br />
    approach mainly consists of two proposals, i.e. deep residual learning for<br />
    image compression and sub-pixel convolution as up-sampling operations.<br />
    Experimental results have indicated that our approaches, Kattolab, Kattolabv2<br />
    and KattolabSSIM, achieve 0.972 in MS-SSIM at the rate constraint of 0.15bpp<br />
    with moderate complexity during the validation phase.

  • Learning image and video compression through spatial-temporal energy compaction

    Zhengxue Cheng, Heming Sun, Masaru Takeuchi, Jiro Katto

    Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition   2019-June   10063 - 10072  2019.06

     View Summary

    Compression has been an important research topic for many decades, to produce a significant impact on data transmission and storage. Recent advances have shown a great potential of learning based image and video compression. Inspired from related works, in this paper, we present an image compression architecture using a convolutional autoencoder, and then generalize image compression to video compression, by adding an interpolation loop into both encoder and decoder sides. Our basic idea is to realize spatial-temporal energy compaction in learning image and video compression. Thereby, we propose to add a spatial energy compaction-based penalty into loss function, to achieve higher image compression performance. Furthermore, based on temporal energy distribution, we propose to select the number of frames in one interpolation loop, adapting to the motion characteristics of video contents. Experimental results demonstrate that our proposed image compression outperforms the latest image compression standard with MS-SSIM quality metric, and provides higher performance compared with state-of-the-art learning compression methods at high bit rates, which benefits from our spatial energy compaction approach. Meanwhile, our proposed video compression approach with temporal energy compaction can significantly outperform MPEG-4, and is competitive with commonly used H.264. Both our image and video compression can produce more visually pleasant results than traditional standards.

    DOI

  • Methods for adaptive video streaming and picture quality assessment to improve QoS/QoE performances

    Kenji Kanai, Bo Wei, Zhengxue Cheng, Masaru Takeuchi, Jiro Katto

    IEICE Transactions on Communications   E102B ( 7 ) 1240 - 1247  2019

     View Summary

    This paper introduces recent trends in video streaming and four methods proposed by the authors for video streaming. Video traffic dominates the Internet as seen in current trends, and new visual contents such as UHD and 360-degree movies are being delivered. MPEG-DASH has become popular for adaptive video streaming, and machine learning techniques are being introduced in several parts of video streaming. Along with these research trends, the authors also tried four methods: route navigation, throughput prediction, image quality assessment, and perceptual video streaming. These methods contribute to improving QoS/QoE performance and reducing power consumption and storage size.

    DOI

  • Throughput Prediction Using Recurrent Neural Network Model

    Bo Wei, Mayuko Okano, Kenji Kanai, Wataru Kawakami, Jiro Katto

    2018 IEEE 7th Global Conference on Consumer Electronics, GCCE 2018     88 - 89  2018.12

     View Summary

    To ensure good quality of experience for user when transmitting video content, throughput prediction can contribute to the selection of proper bitrate. In this paper, we propose a throughput prediction method with recurrent neural network (RNN) model. Experiments are conducted to evaluate the methods, and the results indicate that proposed method can decrease the prediction error by a maximum of 29.39% compared with traditional methods.

    DOI

display all >>

Industrial Property Rights 【 display / non-display

display all >>

Awards 【 display / non-display

  • 電子情報通信学会 フェロー

    2015.09  

  • 電子情報通信学会 通信ソサイエティ 活動功労賞

    2006.09  

  • 電子情報通信学会 学術奨励賞

    1995.09  

  • SPIE VCIP 1991, Best Student Paper Award

    1991.11  

Research Projects 【 display / non-display

  • 無線LANを用いた車載APの広域被覆配置による広域高速大容量通信

    基盤研究(B)

    Project Year :

    2019.04
    -
    2022.03
     

    村瀬 勉, 金井 謙治, 甲藤 二郎, 計 宇生, 策力 木格, 小畑 博靖, 塩田 茂雄, 宮田 純子

     View Summary

    以下の3つの研究テーマを実施した。
    ■研究テーマ1: 車の移動制御・最適ルート誘導制御においては、(a)目的地に向かう車両の最短経路ではなく、需要のある場所への迂回ルートをとったときの、得失を評価する。■成果:車両が移動する経路を2.5倍居ないに納めるという条件のもと、最適なルート制御を行うことで、最短経路の場合よりも、スループットを280%(=約4倍)向上させることができる、という結果を得た。これにより、車両のルート制御をうまく行うことで、所望のスループットを得るという当初の目的が概ね達成できる見込みとなった。
    ■研究テーマ2: 干渉緩和技術・最適AP選択技術においては、(d) APへの接続可否制御、APへの負荷分散制御を提案する。■成果: 移動するAPに連続して接続するための基礎検討を行った。IEEE802.11adのミリ波通信においては、移動する車との通信は、非常に短時間に高速で情報を転送する必要があることがわかった。そのため、現在普及しているIEEE802.11acなどの帯域では、混信(干渉)が大きすぎ実用が困難であることがわかった。一方、11adを用いる本研究の方法では、干渉や混信の影響よりもハンドオーバの制御方法で性能が決まることが判明し、ハンドオーバの最適化方法として、ビーコン送出間隔の調整と、ハンドオーバタイミングを決定する技術を開発した。
    ■研究テーマ3: 高速短時間通信技術においては、(g)IEEE802.11adのミリ波通信方法の移動通信における性能を評価する。■成果:車内のアクセスポイントと車外の歩行者のスマホとで通信するモデルにおいて、ミリ波通信は、車両のボディが障害物とある影響で性能が大きく変わってくることが明らかになった。特に金属ボディでの減衰は極めて大きく、逆に、ガラスを通した場合には、それほどでも無いことなどを定量的に明らかにした。

  • Next generation networking infrastructure and application verification towards the most advanced mobile ICT system

    Grant-in-Aid for Scientific Research (A)

    Project Year :

    2015.04
    -
    2019.03
     

    Katto Jiro

     View Summary

    We set next research & development issues in this project. As core techniques, (1) collection of wireless communication records, (2) prediction of wireless communication quality based on the record history, and (3) delivery control and route navigation based on the quality prediction. As extension techniques, (4) large scale deployment, security, sensor assist, new wireless communication support, and QoE evalutions, and (5) prototype implementations. For (1) to (3), we proposed quality prediction methods using machine learning, adaptive delivery control maximizing QoE metrics, and moving route navigation maximizing communication quality such as throughputs. For (4) to (5), we tried cloud system extension, performance improvement by using additional sensors and implementation experiments over actual networks. Finally, we published our research contributions in international conferences and transactions with peer reviews.

  • User Cooperative Mobility for Communication Quality in Densely Deployed Wireless LANs

    Grant-in-Aid for Scientific Research (B)

    Project Year :

    2015.04
    -
    2018.03
     

    MURASE TUTOMU, ONISHI Hirofumi

     View Summary

    To improve quality of services (QoS) without changing protocols or specifications, this research proposed “user mobility control” in which users or devices move the better place to obtain better QoS. If users can move, then we have more better parameters of protocols fitting to the current user positions. Tradeoff of the mobility cost and QoS improvement was evaluated by analysis, simulations and real device experiments with using newly developed proposed methods and algorithms. For maultiple users movements, heuristic approached are also developed.

  • Development of Software Radio and Audio Platform and Video Delivery Experiment

    Grant-in-Aid for Challenging Exploratory Research

    Project Year :

    2013.04
    -
    2015.03
     

    KATTO Jiro

     View Summary

    In this research, we developed platforms for software radio and software audio and carried out evaluation experiments. For software radio, we used GNU Radio, measured communication characteristics by changing transmission power, modulation and error correction, and tried wireless delivery of music, still images and motion pictures as its application. For software audio, we used commercial speakers, microphones and MATLAB, measured communication characteristics, and tried delivery of music and still images modulated over audio signals. Our software audio failed video transmission due to low data rates, but the results were paid attention and accepted by IEEE WCNC 2015 heled in March, 2015.

  • Performance Improvement and Implementation of Hybrid TCP Congestion Control

    Grant-in-Aid for Scientific Research (B)

    Project Year :

    2008
    -
    2010
     

    KATTO Jiro, SU Zhou

     View Summary

    This research had focused on transport protocols which were adequately designed and applied to broadband wired networks, wireless networks and underwater sensor networks. For each topic, novel proposals which outperformed conventional ones had been made by integrating theoretical analysis, simulation evaluations and actual implementations. We had also achieved future direction toward integrated design of transport protocols.

display all >>

Specific Research 【 display / non-display

  • 深層学習を用いた画像圧縮・画像通信に関する研究開発

    2020  

     View Summary

    本研究開発では、深層学習を用いた画像圧縮と画像通信に関する研究開発を行った。画像圧縮に関しては、動画像圧縮におけるイントラ予測の特性改善、エンドエンド学習型画像圧縮の固定小数点実装、RNNを用いた学習型画像圧縮、フレーム補間を活用した動画像圧縮、等に関する検討を行った。また、成果発表は、IEEE TMM、CVPR、VCIP、ICIP、ICASSP等で行い、VCIPにおける発表はBest Paper Awardを受賞した。画像通信に関しては、映像ストリーミングにおける適応レート制御、無線通信に対する超解像応用、等の検討を行った。また、成果発表は、国際学会のWCNCとCCNCで行った。

  • 深層学習適用による革新的な画像圧縮と映像配信に関する研究開発

    2019  

     View Summary

    本研究開発では、深層学習を用いた画像圧縮と画像通信に関する研究開発を行った。画像圧縮に関しては、深層学習を用いた静止画像の非可逆圧縮と可逆圧縮、動画像の非可逆圧縮に関する検討を行った。画像通信に関しては、深層学習を用いた適応レート制御と360度映像拡張に関する検討を行った。研究成果は、CVPR、ICIP、VCIP、PCS、ISM、ICCE等の国際会議で発表を行うと共に、IEEE Trans. CSVT、IEEE Access、IEICE Trans. Comm. 等の査読付き論文誌にも採録された。また、2020年度も、CVPR、ICASSP等の国際学会やIEEE Trans. Multimedia、IEEE Trans. CSVT等の論文誌への採録が確定している。

  • 遅延クリティカルアプリケーションのための基盤技術開発と体系化

    2017  

     View Summary

    本研究開発では、クラウドを活用するマルチメディアサービスについて、サービス時間の短縮に貢献するエッジコンピューティングのプロトタイプ実装と特性評価を行った。一つの検討例では、OpenStackを活用したエッジコンピューティングのプロトタイプシステムを作成し、具体的なアプリケーションとして映像配信と人物検知を実装した。また、別の検討例では、エッジコンピューティングを活用した映像監視を想定し、人物検知の結果に応じて配信レートを増減する適応映像配信システムを作成した。その上で、それぞれのプロトタイプの実機実験評価として、低遅延化の効果を確認した。

  • RFファインダーの研究開発

    2015  

     View Summary

    無線通信情報と可視光・赤外線カメラの併用によって無線通信機器の位置をピンポイントで特定するセンサフュージョン型測位システム「RFファインダー」の研究開発を進めた。無線LANアクセスポイントを対象に、スマートフォンを用いたSSIDや電波受信電力の取得によっておよその位置を推定し、赤外線カメラによって、熱源として無線LANアクセスポイントの位置を特定した。一方、Raspberry PiなどのIoTデバイスも試したが、こちらは熱源として特定するには、温度がさほど高くならないことも確認した。また、可視光カメラと赤外線カメラをハイブリッド使用した人物検出の精度改善についても実験を行ない、成果発表を行なった。

  • 快適で省電力なスマート・ワイヤレス・ナビゲーション

    2014  

     View Summary

    快適で省電力な無線通信を実現するスマート・ワイヤレス・ナビゲーションに関する研究開発を進めてきた。スマートフォンなどを用いて、セルラー基地局と無線LANアクセスポイントの位置、時刻、通信品質、消費電力を収集し、無線信号マップを作成する。そして、そのマップを活用して通信品質の未来予測を行ったり、通信品質を最大化する経路を提示したりすることに応用する。今年度は科学研究費補助金の採択を目指し、論文投稿と学会発表の実績作りを進め、最終的には基盤研究(A)に採択されることができた。

display all >>

 

Syllabus 【 display / non-display

display all >>

 

Committee Memberships 【 display / non-display

  • 2014.06
    -
    2016.06

    映像情報メディア学会  編集長

  • 2012
    -
    2015

    IEEE ComSoC  Tech News Editorial Board

  • 2012.06
    -
    2014.05

    画像電子学会  財務理事

  • 2007.05
    -
    2013.04

    電子情報通信学会 ネットワークシステム研究専門委員会  専門委員

  • 2013
    -
     

    IEEE Healthcom 2013  Technical Program Committee

display all >>