Recently, aging of the population and medical cost inflation are emerging as social issues to be solved. To address this problem, estimations of bioinformation based on Internet of Things (IoT) and machine learning get more attention to researchers. In this paper, in order to estimate R-R Interval (RRI) without using specialized and professional wearable devices, we propose a deep learning based RRI estimation method by using mainly smartphone sensors. For dataset, we collect ECG (for label), 3-axis acceleration, pressure, illuminance, GPS and temperature (for training data) under different exercise types (walking and running) by using a smartphone and smart clothing called hitoe. To construct a regression model, we adopt a dual stage attention-based RNN model. From the evaluation results, we confirm that the proposed method can estimate RRI and LF/HF with acceptable accuracy.

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.

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.

Convolutional neural networks (CNNs) have achieved great successes on many computer vision tasks, such as image recognition, video processing, and target detection. In recent years, many hardware designs have been devoted to accelerating CNN inference. In order to further speed up CNN inference and reduce data waste, this work proposed a reconfigurable sliced systolic array: 1) Depending on the number of network nodes in each layer, the slice mode could be dynamically configured to achieve high throughput and resource utilization. 2) To take full advantage of convolution reuse and weight reuse, this work designed a tile-column sliding (TCS) processing dataflow. 3) A four-stage for loop algorithm was employed, which divides the CNN calculation into several parts based on the input nodes and output nodes. The entire CNN inference is carried out using integer-only arithmetic originated from TensorLite. Experimental results prove that these strategies lead to significant improvement in inference performance and energy efficiency.

As the demand for high-resolution videos grows, the next generation video coding standard Versatile Video Coding introduces many new proposals, including Adaptive Multiple Transforms (AMT), to improve coding efficiency. This paper presents a reconfigurable transform core for the VVC standard where the implementation of 1D DST-VII and DCT-VIII for all transform sizes are enabled. To offer a very low circuit complexity, a simple approximation strategy with a little coding performance loss is proposed. An 8×8 Processing Element (PE) array is employed as the core computational unit, where each PE can be configured dynamically based on the transform type. In addition, the transforms of larger sizes can be realized in the finite PE units with the Partitioned Matrix Multiplication (PMM) scheme. The experimental and synthesis results show that this design can save at least 29.1% area compared with other works in literature with the negligible degradation of video quality and a slight increase in the bit rate.

With the increasing video demand in daily network traffic, it is an urgent task to develop effective algorithms to facilitate high-quality content delivery service. Recently, numerous adaptive streaming algorithms have been proposed to improve the user perceived experience. However, these algorithms were mainly developed from the perspective of single user. There is not yet systematical evaluation and comparison of the bitrate adaptation methods for multi-user video streaming. Besides, the Quality of Experience (QoE) metrics were not unified. In this work, we propose a new mininet-based testbed framework which is able to conduct real-time video streaming emulation in various multi-user scenarios. Seven state-of-the-art adaptation methods are incorporated into the testbed. Meanwhile, ITU-T P.1203 model, the world's first standard for measuring QoE of HTTP adaptive streaming, is implemented to calculate the mean opinion scores of different methods. Using the developed testbed, the performance of current adaptation methods in multi-user network is analyzed and compared. A variety of experiments are carried out by changing the user number and network conditions, in which the QoE of different users are investigated. It is found that current algorithms perform inconsistently in various network scenarios. In the excessive user and limited bandwidth cases, machine learning and scheduling techniques show superiority in providing high and equal QoE for all users. While in the high-delay case, the buffer-based approaches show robust performance. Overall, the findings of this work give an insight for designing and choosing adaptive streaming strategies in different multi-user network conditions.

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.

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.

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.

Recently, an efficient and automatic infrastructure maintenance service is mandatory. To address this demand, in this paper, we introduce a segmentation-based road damage detection method by using U-Net. To train the model, we collect 4K images by using a smartphone mounted on a bicycle and build our own road damage dataset. In addition, to improve detection accuracy, we apply focal loss and image patch for loss function and input image, respectively. From the evaluation, the result confirms that the method demonstrates to extract road damages with acceptable accuracy.

In this paper, we propose a deep learning-based distance estimation method from a single-shot image. In the proposal, we model the estimation as a regression problem, and estimate the distance between a pedestrian and a camera by using three main features; size of bounding box, image blur and image features. By using KITTI dataset, we evaluate the accuracy of the proposed model.

Recently, in order to build barrier-free smart cities, Japanese government is actively promoting to construct open data regarding roads and buildings, including existence of bumps, road slope and corridor width. To contribute this activity, in this paper, we introduce a corridor width estimation method by using an RGB-D camera. In this method, we assume that the RGB-D camera is mounted on the electric wheelchairs for a mobile sensing platform. The corridor width estimation is mainly operated by three steps: planes extraction, side walls extraction and distance calculation. We evaluate the accuracy of this method at eight corridors in Waseda University.

To contribute to publishing pedestrian networks as open data, this paper introduces a curb height estimation method by using 2DLiDAR mounted on e-Bike. We develop a prototype system by using e-Bike, 2DLiDAR and other sensors, and collect actual curb data around Waseda University. From accuracy evaluations, the results confirm that the prototype can estimate road level differences and positions with high accuracy.

Under complex network conditions, adaptive video streaming requires additional state information for optimal quality selection. In this paper, we present the applicability of reinforcement learning techniques on NDN adaptive streaming. Both buffer-based and throughput-based adaptation are studied and observed their characteristics. The Q-learning algorithm is used to learn state-action values. Based on a greedy policy, the simulation results demonstrate that RL agents tend to choose the best possible bitrate which consequently reduces the quality fluctuation in adaptive streaming.

Learned image compression (LIC) has reached the traditional hand-crafted methods such as JPEG2000 and BPG in terms of the coding gain. However, the large model size of the network prohibits the usage of LIC on resource-limited embedded systems. This paper presents a LIC with 8-bit fixed-point weights. First, we quantize the weights in groups and propose a non-linear memory-free codebook. Second, we explore the optimal grouping and quantization scheme. Finally, we develop a novel weight clipping fine tuning scheme. Experimental results illustrate that the coding loss caused by the quantization is small, while around 75% model size can be reduced compared with the 32-bit floating-point anchor. As far as we know, this is the first work to explore and evaluate the LIC fully with fixed-point weights, and our proposed quantized LIC is able to outperform BPG in terms of MS-SSIM.

Recently learned image compression has achieved many great progresses, such as representative hyperprior and its variants based on convolutional neural networks (CNNs). However, CNNs are not fit for scalable coding and multiple models need to be trained separately to achieve variable rates. In this paper, we incorporate differentiable quantization and accurate entropy models into recurrent neural networks (RNNs) architectures to achieve a scalable learned image compression. First, we present an RNN architecture with quantization and entropy coding. To realize the scalable coding, we allocate the bits to multiple layers, by adjusting the layer-wise lambda values in Lagrangian multiplier-based rate-distortion optimization function. Second, we add an RNN-based hyperprior to improve the accuracy of entropy models for multiple-layer residual representations. Experimental results demonstrate that our performance can be comparable with recent CNN-based hyperprior methods on Kodak dataset. Besides, our method is a scalable and flexible coding approach, to achieve multiple rates using one single model, which is very appealing.

For the next-generation video coding standard Versatile Video Coding (VVC), several new contributions have been proposed to improve the coding efficiency, especially in the transformation operations. This paper proposes a unified 32× 32 block-based transform architecture for the VVC standard that enables 2D Discrete Sine Transform-VII (DST-VII) and Discrete Cosine Transform-VIII (DCT-VIII) of all sizes. It mainly gives three contributions: 1) The N-Dimensional Reduced Adder Graph (RAG-n) algorithm is adopted to design the minimal adder-oriented computational units. 2) The storage of the asymmetric transform units can be realized in the dual-port SRAM-based transpose memory. 3) The pipelined 2D transformations of mixed block sizes are achieved with the throughput rate of 32 samples per cycle. The synthesis results indicate that this architecture can reduce area by up to 73.1% compared with other state-of-the-art works. Moreover, power saving ranging from 4.9% to 9.9% can be achieved. Regarding the transpose memory, at least 21.9% of the area can be saved by using SRAM.

With the continuous development of mobile devices, it becomes possible for people to demand higher definition videos. To alleviate the pressure of deploying the video codec in mobile multimedia, a learning-based low complexity in-loop filter is proposed in this paper. Depthwise separable convolution is combined with batch normalization to construct this model. To enhance its performance, the knowledge from a pre-trained teacher model is transferred to it. However, the over-smoothing problem in the inter frames caused by double enhancing effect remains. To solve this, a Wiener-based filtering algorithm that tries to restore the distortion from the learned residuals is designed and introduces an adequate filtering effect. The experimental results show that our proposed methods achieve considerable BD-rate reduction than HEVC anchor. Compared with the previous state-of-the-art work VR-CNN, our model achieves 1.65% extra BD-rate reduction, 79.1% decrease in FLOPs, 25% decrease in encoding complexity, and 70% decoding complexity decrease.

In this paper, a coding framework VIP-ICT-Codec is introduced. Our method is based on the VTM (Versatile Video Coding Test Model). First, we propose a color space conversion from RGB to YUV domain by using a PCA-like operation. A method for the PCA mean calculation is proposed to de-correlate the residual components of YUV channels. Besides, the correlation of UV components is compensated considering that they share the same coding tree in VVC. We also learn a residual mapping to alleviate the over-filtered and under-filtered problem of specific images. Finally, we regard the rate control as an unconstraint Lagrangian problem to reach the target bpp. The results show that we achieve 32.625dB at the validation phase.

In this paper, we provide a detailed description on our submitted method Kattolab to Workshop and Challenge on Learned Image Compression (CLIC) 2020. Our method mainly incorporates discretized Gaussian Mixture Likeli-hoods to previous state-of-the-art learned compression algorithms. Besides, we also describes the acceleration strategies and bit optimization with the low-rate constraint. Experimental results have demonstrated that our approach Kattolab achieves 0.9761 in terms of MS-SSIM at the rate constraint of 0.15 bpp during the validation phase.

Lossless image compression is an important task in the field of multimedia communication. Traditional image codecs typically support lossless mode, such as WebP, JPEG2000, FLIF. Recently, deep learning based approaches have started to show the potential at this point. HyperPrior is an effective technique proposed for lossy image compression. This paper generalizes the hyperprior from lossy model to lossless compression, and proposes a L2-norm term into the loss function to speed up training procedure. Besides, this paper also investigated different parameterized models for latent codes, and propose to use Gaussian mixture likelihoods to achieve adaptive and flexible context models. Experimental results validate our method can outperform existing deep learning based lossless compression, and outperform the JPEG2000 and WebP for JPG images.

Image compression has been an important research topic for many decades. Recently, deep learning has achieved great success in many computer vision tasks, and its use in image compression has gradually been increasing. In this paper, we present an energy compaction-based image compression architecture using a convolutional autoencoder (CAE) to achieve high coding efficiency. Our main contributions include three aspects: 1) we propose a CAE architecture for image compression by decomposing it into several down(up)sampling operations; 2) for our CAE architecture, we offer a mathematical analysis on the energy compaction property and we are the first work to propose a normalized coding gain metric in neural networks, which can act as a measurement of compression capability; 3) based on the coding gain metric, we propose an energy compaction-based bit allocation method, which adds a regularizer to the loss function during the training stage to help the CAE maximize the coding gain and achieve high compression efficiency. The experimental results demonstrate our proposed method outperforms BPG (HEVC-intra), in terms of the MS-SSIM quality metric. Additionally, we achieve better performance in comparison with existing bit allocation methods, and provide higher coding efficiency compared with state-of-the-art learning compression methods at high bit rates.

The amount of video data is so large that efficient video compression is required for the storage and transmission. Intra prediction is one of important components in video compression. In this paper, we examine various Convolutional Neural Network (CNN) structures to estimate optimal intra prediction mode as Most Probable Modes (MPMs). Moreover, we investigate several combinations of the MPMs obtained by the CNN and MPMs derived from High Efficiency Video Coding Test Model (HM). From these experimental results, we find that using 6 MPMs from both CNN and HM with moderate number of channels or kernel size is preferred.

Different from the traditional quaternary tree (QT) structure utilized in the previous generation video coding standard H.265/HEVC, a brand new partition structure named quadtree with nested multi-type tree (QTMT) is applied in the latest codec H.266/VVC. The introduction of QTMT brings in superior encoding performance at the cost of great time-consuming. Therefore, a fast intra partition algorithm based on variance and Sobel operator is proposed in this paper. The proposed method settles the novel asymmetrical partition issue in VVC by well balancing the reduction of computational complexity and the loss of encoding quality. To be more concrete, we first terminate further splitting of a coding unit (CU) when the texture of it is judged as smooth. Then, we use Sobel operator to extract gradient features to decide whether to split this CU by QT, thus terminating further MT partitions. Finally, a completely novel method to choose only one partition from five QTMT partitions is applied. Obviously, homogeneous area tends to use a larger CU as a whole to do prediction while CUs with complicated texture are prone to be divided into small sub-CUs and these sub-CUs usually have different textures from each other. We calculate the variance of variance of each sub-CU to decide which partition will distinguish the sub-textures best. Our method is embedded into the latest VVC official reference software VTM-7.0. Comparing to anchor VTM-7.0, our method saves the encoding time by 49.27% on average at the cost of only 1.63% BDBR increase. As a traditional scheme based on variance and gradient to decrease the computational complexity in VVC intra coding, our method outperforms other relative existing state-of-the-art methods, including traditional machine learning and convolution neural network methods.

Image compression is a fundamental research field and many well-known compression standards have been developed for many decades. Recently, learned compression methods exhibit a fast development trend with promising results. However, there is still a performance gap between learned compression algorithms and reigning compression standards, especially in terms of widely used PSNR metric. In this paper, we explore the remaining redundancy of recent learned compression algorithms. We have found accurate entropy models for rate estimation largely affect the optimization of network parameters and thus affect the rate-distortion performance. Therefore, in this paper, we propose to use discretized Gaussian Mixture Likelihoods to parameterize the distributions of latent codes, which can achieve a more accurate and flexible entropy model. Besides, we take advantage of recent attention modules and incorporate them into network architecture to enhance the performance. Experimental results demonstrate our proposed method achieves a state-of-the-art performance compared to existing learned compression methods on both Kodak and high-resolution datasets. To our knowledge our approach is the first work to achieve comparable performance with latest compression standard Versatile Video Coding (VVC) regarding PSNR. More importantly, our approach generates more visually pleasant results when optimized by MS-SSIM.

Surveillance cameras are commonly used for security purpose. However, most of them are used for verification after incidents happen. In this paper, we propose a method for proactive video surveillance system using human pose estimation. Our method estimates anomaly scores of human actions using the video descriptors (human pose and bounding box) extracted by pose estimation and tracking methods. We estimate human poses, apply PCA, estimate GMM parameters, and finally calculate anomaly scores based on the GMM. We evaluate our method and compare with higher-level recognition-based method. Experimental results demonstrate effectiveness of human pose for video anomaly detection.

Recently, deep learning techniques have contributed to solving a multitude of computer vision tasks. In this paper, we propose a deep-learning approach for video alignment, which involves finding the best correspondences between two overlapping videos. We formulate the video alignment task as a variant of the well-known machine comprehension (MC) task in natural language processing. While MC answers a question about a given paragraph, our technique determines the most relevant frame sequence in the context video to the query video. This is done by representing the individual frames of the two videos by highly discriminative and compact descriptors. Next, the descriptors are fed into a multi-stage network that is able, with the help of the bidirectional attention flow mechanism, to represent the context video at various granularity levels besides estimating the query-aware context part. The proposed model was trained on 10k video-pairs collected from 'YouTube'. The obtained results show that our model outperforms all known state of the art techniques by a considerable margin, confirming its efficacy.

Quadtree with nested multi-Type tree (QTMT) partition structure in Versatile Video Coding (VVC) contributes to superior encoding performance compared to the basic quad-Tree (QT) structure in High Efficiency Video Coding (HEVC). However, the improvement of performance leads to an un-Avoidable increase of computational complexity. To achieve a balance between coding efficiency and compression quality, we propose a fast intra partition algorithm based on variance and gradient to solve the rectangular partition problem in VVC. First, further splitting of smooth areas is terminated. Then, QT partition is chosen depending on the gradient features extracted by Sobel operator. Finally, one partition from five possible QTMT partitions is directly chosen by computing the variance of variance of sub-CUs. The theoretical basis of our method is that a homogeneous area tends to be predicted with a larger coding unit (CU), and sub-parts of a split CU are prone to have different textures from each other. To our knowledge, this is the first attempt to apply traditional method to accelerating the rectangular partition problem in VVC intra prediction. Experimental results show that the proposed method can save averagely 53.17% encoding time with only 1.62% BDBR increase and 0.09dB BDPSNR loss compared to anchor VTM4.0.

Recently, the demand for high quality 360-degree video delivery is increasing, however, 360-degree videos require extremely high bitrate and large network capacity. Therefore, an efficient (i.e., higher quality and lower traffic) 360-degree video delivery is mandatory. To address this fact, this paper introduces and evaluates a tile-based 360-degree video delivery system that equips viewport movement prediction and rate adaptation.

In this paper, a dual learning-based method in intra coding is introduced for PCS Grand Challenge. This method is mainly composed of two parts: intra prediction and reconstruction filtering. They use different network structures, the neural network-based intra prediction uses the full-connected network to predict the block while the neural network-based reconstruction filtering utilizes the convolutional networks. Different with the previous filtering works, we use a network with more powerful feature extraction capabilities in our reconstruction filtering network. And the filtering unit is the block-level so as to achieve a more accurate filtering compensation. To our best knowledge, among all the learning-based methods, this is the first attempt to combine two different networks in one application, and we achieve the state-of-the-art performance for AI configuration on the HEVC Test sequences. The experimental result shows that our method leads to significant BD-rate saving for provided 8 sequences compared to HM-16.20 baseline (average 10.24% and 3.57% bitrate reductions for all-intra and random-access coding, respectively). For HEVC test sequences, our model also achieved a 9.70% BD-rate saving compared to HM-16.20 baseline for allintra configuration.

In this paper, we present a road infrastructure monitoring system using e-bikes to support smart community. Smartphone and IoT (Internet of Things) devices are attached to e-bikes, and electric power is supplied from the batteries of e-bikes. Image processing techniques based on deep learning are applied and run on IoT devices or cloud backbone. A prototype system is implemented, which performs from image capturing to web browsing of the processed images.

Adaptive streaming on Named Data Networking appears to be a challenging task since the provider is unknown to the consumer. In this paper, we present an approach to efficiently adjust the streaming bit-rate by deploying IoT devices. Each NDN node is designed to send IoT devices the information of available bandwidth. The IoT devices then assist the consumers if the current network can deliver 4K video segments. The experiments using NDNx and NDN-JS show that the consumer in the proposed approach could select representation bit-rate more accurately than the traditional rate adaptation method.

Recently, Ultra HDTV (UHDTV) services become popular over satellite and on the internet. On the contrary, there are tremendously huge volume of High Definition Television (HDTV) and Standard Definition Television (SDTV) contents stored in broadcasting companies and storage devices. In this paper, we propose a color space conversion (also known as gamut mapping) method from BT. 709 (used for current HDTV broadcast) to BT. 2020 (used for UHDTV broadcast), which estimates and restores lost color information. It learns an end-to-end conversion method from BT. 709 image to BT. 2020 image with restoring lost color information using Convolutional Neural Network (CNN). By experiments, we confirm that our method can achieve 2.31dB gain against the conventional method on average.

Intelligent sensing is an important application field of video coding. For the next generation video coding standard Versatile Video Coding (VVC), several new contributions have been proposed to improve the coding efficiency, especially in the transformation operations. This paper proposes a unified transform architecture for VVC standard that enables 1D Discrete Sine Transform-VII (DST-VII) and Discrete Cosine Transform-VIII (DCT-VIII) of all sizes. In order to minimize the number of adders, the N-Dimensional Reduced Adder Graph (RAG-n) algorithm is adopted to design the logical computation of the VVC transforms. In addition, a pipeline operation is employed to achieve a high throughput rate of 32 samples per cycle. The experimental and synthesis results show that our proposal can save 16.96% area and 9.89% power consumption than the direct implementation by the multipliers. Moreover, this design can reduce the normalized area (NA) by at least 57.53% compared with other advanced works.

Abstract-Virtualized service and network functions are deployed on virtual machines (VMs) to realize essential processing to realize service function chaining (SFC). Issues on SFC is SF allocation to a VM and to minimize the response time and number of function instances. In this paper, we propose an SF clustering-based scheduling algorithm, called 'SF-clustering for utilizing virtual CPUs' (SFCUV), to solve the SF allocation and SF selection problems simultaneously. Experimental results show that SF-CUV can utilize vCPUs to minimize the response time.

To develop and interoperate smart city applications efficiently, smart city IoT platforms require efficient handling of various types of sensor devices, networking and computing resources, and different domain applications. To address this fact, the authors of this paper proposed a research project named 'Fed4IoT', which is an acronym for federation of IoT and cloud infrastructures, to provide scalable and interoperable smart city applications. The project primarily proposes two key technologies: IoT device virtualization and context-information sharing. In this paper, we introduce the IoT device virtualization (one of key technology in Fed4IoT) that enables efficient utilization of networking and computing resources by applying a micro-service sharing and dynamic resource scaling. In the performance evaluations, we implement three micro services, such as object detection, garbage detection and road damage detection, as smart city applications and dynamically and flexibly allocate computing resources to three micro services by using Docker and Kubernetes. Through the evaluations we validate that the proposed IoT device virtualization achieves the efficient computing resource utilization and reduction of networking resources.

Recently, the demand for immersive VR (360-degree) video delivery is increasing, and an efficient (high quality and low traffic) 360-degree video streaming methodology is mandatory. To address this fact, in this paper, we introduce a tile-based 360-degree DASH streaming with viewport prediction. We evaluate the prediction accuracy of future viewport movement patterns by using clustering and image and audio information. In addition, we perform rate adaptation based on viewport prediction results. Through evaluations, we confirm the tile-based 360-degree DASH streaming achieves higher objective image quality and lower total video traffic volume.

Towards an intelligent video surveillance system (high throughput and high detection accuracy), in this paper, we quantitatively validate the network performance of 60GHz IEEE 802.11ad and the contribution of 4K resolution to image-based human pose detection. In the performance evaluation of 11ad, we evaluate TCP throughputs by changing the communication distances (up-to 6m). In the performance evaluation of human pose detection, we evaluate the detection accuracy of human pose detection by changing a distance between a camera and a subject (up-to 100 m) and input image resolutions.

To realize highly reliable video surveillance and provide ultrahigh-definition/immersive video streaming, it is planned to adopt the 5G cellular system using millimeter-wave (mmWave) as the wireless-network infrastructure. However, mmWave communication has a challenging issue: mmWave communication is extremely sensitive to obstacles, such as walls, pillars, and even human bodies, and this issue easily increases the packet loss rates and round trip time (RTT) (or disconnection from the base station) due to a no line of sight (NLOS) environment. Therefore, in this work, 5G throughput performances were evaluates in an indoor train station by considering the effect of an NLOS environment caused by blockage by human bodies. In addition, to improve the robustness of TCP transmission in a high-RTT and high-packet-loss environment (e.g., an NLOS environment), a state-of-the-art TCP, TCP-FSO, was used. In the evaluations, the MATLAB 5G library was used to simulate the 5G environment, and a Linux software-based network emulator, Traffic Control, was used to emulate the 5G network. From the evaluations, it the 5G mobile throughput characteristics were confirmed in three different crowded patterns (low, middle, and high density), and the TCP-FSO advantage against CUBIC-TCP was validated.

To recognize transportation modes without any additional sensor devices, we demonstrate that the transportation modes can be recognized from communication quality factors. In the demonstration, instead of using global positioning system (GPS) and accelerometer sensors, we collect mobile TCP throughputs, received-signal strength indicators (RSSIs), and cellular base-station IDs (Cell IDs) through in-line network measurement when the user enjoys mobile services, such as video streaming. In accuracy evaluations, we conduct two different field experiments to collect the data in six typical transportation modes (static, walking, riding a bicycle, riding a bus, riding a train and riding a subway), and then construct the classifiers by applying a support-vector machine (SVM), k-nearest neighbor (k-NN), random forest (RF), and convolutional neural network (CNN). Our results show that these transportation modes can be recognized with high accuracy by using communication quality factors as well as the use of accelerometer sensors.

New video coding standards bring high coding efficiency at the cost of large computational complexity. Especially, in 8K sequence, computation cost becomes much larger due to its super high spatial resolution. In this paper, we propose a method to improve 8K UHDTV coding performance of H.265/HEVC. Our proposed method pre-analyzes spatial and motion complexity, and selects an optimal GOP size or intra/inter prediction mode. Experimental results suggest that the proposed method can improve coding efficiency and reduce computational cost.

The popularity of high dynamic range (HDR) makes the inverse tone mapping become an important technique for HDR display. In this paper, we propose a convolutional neural network (CNN)-based inverse tone mapping method to generate a high-quality HDR image from one single standard dynamic range (SDR) image. First, we present a CNN design with a three- channel input, which considers both luminance and chrominance. Second, we propose to use overlapped inputs to remove the boundary artifacts, caused by zero paddings in CNN. Experimental results demonstrate the high quality of our generated HDR images compared to the ground truth.

Recently, with the spread of Internet of Things (IoT), video surveillance system is widely used to detect abnormal activity using cameras and sensors. In such video surveillance system, contact-free heart rate measurement is a highly required technology to predict the occurrence of abnormal activities and suspicious humans. In this paper, our main contributions can be summarized by two aspects. First, we compare the accuracy performance of heart rate measurement using six heart beat waveform acquisition methods and two heart rate calculation methods. Second, we discuss the influence of image resolutions of RGB cameras on accurate performance, to investigate the requirement of surveillance cameras in actual scenarios.

To develop and interoperate smart city applications efficiently, smart city IoT platforms require efficient handling of various types of sensor devices, networking and computing resources, and different domain applications. To address this fact, in this paper, we introduce an IoT device virtualization that enables efficient utilization of computing resources. The proposal applies a micro-service sharing and dynamic resource scaling. In the performance validations, we implement an early prototype using Docker, Kubernetes, and Apache Kafka. Through the preliminary experiment, we confirm that the proposal can improve the application processing time by appropriately sharing and scaling micro services.

In video streaming on the internet, effective encoding recipes (i.e. bitrate-resolution pairs) are a main obstacle to deliver high-quality video streams. We developed a method to generate an encoding recipe that considers subjective visual quality with one just-noticeable difference (JND) distance. However, this method requires excessive computation time, which is not directly applicable for live streaming. In this paper, in order to provide a light-weight method for live streaming, we developed three acceleration techniques: resolution extrapolation, VMAF skipping and sampled objective measure calculation. These techniques are heuristic, but greatly contribute to reducing computational cost. Experimental results demonstrate that the proposed method achieves a significant reduction in computation time without significant effects on rate-JND characteristics.

Recently, ultra high definition television (UHDTV) services have become popular using satellites and the Internet. However, there are expansive volumes of high definition television (HDTV) and standard definition television (SDTV) contents held by broadcasting companies and in storage devices. Herein we propose two color space conversion (also known as gamut mapping) methods from BT.709 (used for current HDTV broadcast) to BT.2020 (used for UHDTV broadcast) that restore or estimate lost color information. One of our methods anisotropically diffuses the BT.709 chromaticities with regard to the direction of the original chromaticities in the BT.2020 color space, generating chromaticities out of BT.709 gamut. The other learns an end-to-end conversion method from a BT.709 image to a BT.2020 image and restores lost color information using convolutional neural network (CNN). Using these methods along with BT.709 images, we obtain BT.2020 images with chromaticities from the BT.709 color gamut.

Dynamic adaptive video streaming over HTTP (DASH) is widely studied and has been adopted in modern video players to ensure user quality of experience (QoE). In DASH, adaptive bitrate control is a key part whose ultimate goal is to maximize video bitrate while minimizing rebuffering. Throughput prediction plays an important role in helping select the proper video bitrate dynamically. In this paper, we studied the influence of throughput prediction on adaptive video streaming. Because the real-world network is dynamic, different methods need to be tested with large-scale deployments and analyzed statistically. However, this is difficult in academic research. Therefore, we established a reproducible trace-based emulation environment, which enables us to compare different methods quantitatively under the artificially same condition, with limited experiments. The throughput prediction methods are implemented into DASH to evaluate the effect on QoE for video streaming. The results indicate that the prediction method using long short-term memory (LSTM) performs better than the other methods. However, throughput prediction alone is not enough to ensure high QoE. To further improve the QoE, we proposed the decision map method (DMM), where the buffer occupancy is also incorporated to make a selection. By using this decision map, the choice of bitrate can be smarter than that when only prediction information is used. The total QoE is further improved by 32.1% in the ferry trace, which shows the effectiveness of DMM in further improving the performance of throughput prediction in adaptive bitrate control.

In this paper, a novel technique is introduced to address the video alignment task which is one of the hot topics in computer vision. Specifically, we aim at finding the best possible correspondences between two overlapping videos without the restrictions imposed by previous techniques. The novelty of this work is that the video alignment problem is solved by drawing an analogy between it and the machine comprehension (MC) task in natural language processing (NLP). Simply, MC seeks to give the best answer to a question about a given paragraph. In our work, one of the two videos is considered as a query, while the other as a context. First, a pre-trained CNN is used to obtain high-level features from the frames of both the query and context videos. Then, the bidirectional attention flow mechanism; that has achieved considerable success in MC; is used to compute the query-context interactions in order to find the best mapping between the two input videos. The proposed model has been trained using 10k of collected video pairs from”YouTube”. The initial experimental results show that it is a promising solution for the video alignment task when compared to the state of the art techniques.

Named Data Networking (NDN) has been recently introduced as a new network architecture. The implementation of adaptive bit-rate video streaming on NDN is one of the most challenging tasks. The in-network caching feature of NDN could mislead a consumer to overestimate the end-to-end throughput for the next video segment. In this paper, we propose a strategy for more efficient rate adaptation by allowing Interest message adaptation in a hop-by-hop fashion. Moreover, the proactive content caching is designed to be utilized on the related router when a bottleneck on the transmission path was detected. The experiments using NDN-JS and DASH-JS demonstrate that our proposed solution provides better average stream bit-rate than the traditional rate adaptation method. Also, the link bandwidth consumption for video data transmission using the proposed scheme could be reduced efficiently.

We introduce a perceptual video quality driven video encoding solution for optimized adaptive streaming. By using multiple bitrate/resolution encoding like MPEG-DASH, video streaming services can deliver the best video stream to a client, under the conditions of the client's available bandwidth and viewing device capability. However, conventional fixed encoding recipes (i.e., resolution-bitrate pairs) suffer from many problems, such as improper resolution selection and stream redundancy. To avoid these problems, we propose a novel video coding method, which generates multiple representations with constant JustNoticeable Difference (JND) interval. For this purpose, we developed a JND scale estimator using Support Vector Regression (SVR), and designed a pre-encoder which outputs an encoding recipe with constant JND interval in an adaptive manner to input video.

Image compression has been investigated as a fundamental research topic for many decades. Recently, deep learning has achieved great success in many computer vision tasks, and is gradually being used in image compression. In this paper, we present a lossy image compression architecture, which utilizes the advantages of convolutional autoencoder (CAE) to achieve a high coding efficiency. First, we design a novel CAE architecture to replace the conventional transforms and train this CAE using a rate-distortion loss function. Second, to generate a more energy-compact representation, we utilize the principal components analysis (PCA) to rotate the feature maps produced by the CAE, and then apply the quantization and entropy coder to generate the codes. Experimental results demonstrate that our method outperforms traditional image coding algorithms, by achieving a 13.7% BD-rate decrement on the Kodak database images compared to JPEG2000. Besides, our method maintains a moderate complexity similar to JPEG2000.

Image quality assessment (IQA) is an inherent problem in the field of image processing. Recently, deep learning-based image quality assessment has attracted increased attention, owing to its high prediction accuracy. In this paper, we propose a fully-blind and fast image quality predictor (FFIQP) using convolutional neural networks including two strategies. First, we propose a distortion clustering strategy based on the distribution function of intermediate-layer results in the convolutional neural network (CNN) to make IQA fully blind. Second, by analyzing the relationship between image saliency information and CNN prediction error, we utilize a pre-saliency map to skip the non-salient patches for IQA acceleration. Experimental results verify that our method can achieve the high accuracy (0.978) with subjective quality scores, outperforming existing IQA methods. Moreover, the proposed method is highly computationally appealing, achieving flexible complexity performance by assigning different thresholds in the saliency map.

Because mobile users demand a high quality and energyfriendly video delivery service that efficiently uses wireless resources, we introduce an energy-efficient video delivery system by applying moving route navigation and playout buffer control based on the mobile throughput history data. The proposed system first determines the optimal travel route to achieve high-speed and energy-efficient communications. Then when a user enters a high throughput area, our system temporarily extends the video playout buffer size, and the user aggressively downloads video segments via a high-speed and energy-efficient wireless connection until the extended buffer is filled. After leaving this area, the user consumes video segments from the extended buffer in order to keep smooth video playback without wireless communications. We carry out computer simulations, laboratory and field experiments and confirm that the proposed system can achieve energy-efficient mobile video delivery.

Throughput prediction is one of the promising techniques to improve the quality of service (QoS) and quality of experience (QoE) of mobile applications. To address the problem of predicting future throughput distribution accurately during the whole session, which can exhibit large throughput fluctuations in different scenarios (especially scenarios of moving user), we propose a history-based throughput prediction method that utilizes time series analysis and machine learning techniques for mobile network communication. This method is called the Hybrid Prediction with the Autoregressive Model and Hidden Markov Model (HOAH). Different from existing methods, HOAH uses Support Vector Machine (SVM) to classify the throughput transition into two classes, and predicts the transmission control protocol (TCP) throughput by switching between the Autoregressive Model (AR Model) and the Gaussian Mixture Model-Hidden Markov Model (GMM-HMM). We conduct field experiments to evaluate the proposed method in seven different scenarios. The results show that HOAH can predict future throughput effectively and decreases the prediction error by a maximum of 55.95% compared with other methods.

Image compression has been investigated for many decades. Recently, deep<br />
learning approaches have achieved a great success in many computer vision<br />
tasks, and are gradually used in image compression. In this paper, we develop<br />
three overall compression architectures based on convolutional autoencoders<br />
(CAEs), generative adversarial networks (GANs) as well as super-resolution<br />
(SR), and present a comprehensive performance comparison. According to<br />
experimental results, CAEs achieve better coding efficiency than JPEG by<br />
extracting compact features. GANs show potential advantages on large<br />
compression ratio and high subjective quality reconstruction. Super-resolution<br />
achieves the best rate-distortion (RD) performance among them, which is<br />
comparable to BPG.

In recent years, the system using acoustic communication is increasing. However, because acoustic communication uses low frequency, transmission rate is lower than radio wave communication. In wireless communication, MIMO-OFDM is proposed for improvement quality and transmission rate. In this paper, we introduce a software-defined acoustic communication platform by using MATLAB and implement acoustic MIMO-OFDM transmission into the platform. Also, we evaluate BER characteristics in various experimental parameters in MATLAB simulation and real environment. Moreover, we evaluate image quality in actual acoustic image transmission by using the acoustic communication platform and we can successfully transmit the image via acoustic MIMO-OFDM.

Image compression has been investigated as a fundamental research topic for<br />
many decades. Recently, deep learning has achieved great success in many<br />
computer vision tasks, and is gradually being used in image compression. In<br />
this paper, we present a lossy image compression architecture, which utilizes<br />
the advantages of convolutional autoencoder (CAE) to achieve a high coding<br />
efficiency. First, we design a novel CAE architecture to replace the<br />
conventional transforms and train this CAE using a rate-distortion loss<br />
function. Second, to generate a more energy-compact representation, we utilize<br />
the principal components analysis (PCA) to rotate the feature maps produced by<br />
the CAE, and then apply the quantization and entropy coder to generate the<br />
codes. Experimental results demonstrate that our method outperforms traditional<br />
image coding algorithms, by achieving a 13.7% BD-rate decrement on the Kodak<br />
database images compared to JPEG2000. Besides, our method maintains a moderate<br />
complexity similar to JPEG2000.

This paper proposes a transmission control protocol (TCP) for long distance high-speed wireless communications, including free-space optical communications (FSOC). Extreme high frequency of wireless com- munications enables high-speed bit rate, but frequent signal error, including burst error, can be a quite severe problem for ordinary high-speed TCPs. To achieve 10 Gbps or higher data transfer throughput on FSOC, the proposed TCP (designated "TCP-FSO") has improved and new features including multi-layer congestion control, retransmission control with packet loss point estimation, delay-based ACK congestion control, and ACK retransmission control. We evaluated data transfer throughput of TCP-FSO and the other TCPs, by throughput model analysis and experiment on real implementa-Tion. Obtained results show that TCP-FSO achieves far higher data transfer throughput than other high-speed TCPs. For example, it achieved a thou- sand times higher throughput than the other high-speed TCPs in a real FSOC environment.

In this paper, we consider a color space conversion method from BT.709 used for current HDTV (High Definition Television) broadcast to BT.2020 that will be used for UHDTV (Ultra HDTV) broadcast, with lost color information restoration. Our method anisotropically diffuses BT.709 chromaticities considering the direction to the original chromaticities in BT.2020 color space. With this method and BT.709 images, we obtained BT.2020 images that had chromaticities out of BT.709 color gamut.

For 8K UHDTV video, which has high spatial resolution, CU size has large influence to encoding efficiency. In addition, a moving object with random motion has also a large impact on encoding efficiency of inter prediction. In this paper, we therefore propose a method to improve H.265/HEVC encoding efficiency by CU size expansion and inter/intra prediction mode selection. Moreover, we evaluated encoding efficiency and complexity by chancing max CU sizes and inter/intra prediction modes. From these evaluation results, the proposed method achieved high efficiency and low computational cost.

As mobile multimedia services have significantly evolved and diversified with the spread of smartphones and Internet of Things (IoT) devices, low-delay multimedia cloud computing is the need of the hour. To address this demand, in this study, we introduce an edge cloud system that equips a multimedia service function chaining capability. A prototype implementation of the proposed edge cloud system has three main features: 1) edge computing deployment by using OpenStack, 2) multimedia service slicing and chaining, and 3) efficient resource management in edge networks. Based on these features, the proposed system achieves lower multimedia processing delay compared to a conventional cloud computing platform. We deploy the proposed system in our laboratory and validate the system performance by using typical multimedia application, such as human detection in video surveillance.

To provide energy-efficient (i.e., longer lifetime of sensors) and network-friendly (i.e., reducing network traffic) field sensing, we propose an edge-centric field monitoring system which applies efficient sensors and camera control. The proposed system detects conditions in a monitoring area and controls sensing frequency (sampling rate) of sensors, and capture rate and encoding rate of surveillance cameras, according to the detected conditions. In addition, the system applies a Multi-access Edge Computing (MEC) platform to provide fast feedback control to the sensors and cameras. In performance evaluations, we assume that the monitoring target is landslide detection and create a miniature 'artificial landslide generation' environment in our laboratory. By using the environment, we evaluate the system performance, and evaluation results indicate that the proposed system can reduce network traffic and save energy consumption efficiently.

To reduce the backbone video traffic generated by video surveillance, we propose an intelligent video surveillance system that offers multi-modal sensor-based event detection and event-driven video rate adaptation. Our proposed system can detect pedestrian existence and movements in the monitoring area by using multi-modal sensors (camera, laser scanner and infrared distance sensor) and control surveillance video quality according to the detected events. We evaluate event detection accuracy and video traffic volume in the experiment scenarios where up to six pedestrians pass through and/or stop at the monitoring area. Evaluation results conclude that our system can significantly reduce video traffic while ensuring high-quality surveillance.

Recently, to provide a low-latency mobile computing platform, Mobile Edge Computing (MEC) is proposed. In this paper, we first summarized the feature capabilities of MEC, such as content distribution and caching, computational offloading and multimedia Internet of Things (IoT). Then, to understand recent research efforts of multimedia MEC, we briefly highlight the research efforts in terms of above three capabilities: how to achieve edge caching in video distribution, how to schedule computational offloading to the cloud and how the communication quality degradation affects to the user experience of multimedia IoT. Finally, we addressed the emerging research issues of multimedia MEC to improve reliability and robustness of multimedia MEC.

Throughput prediction is one of good solutions to improve quality of mobile applications (e.g., YouTube or Netflix) for video streaming delivery services in mobile networks. This is because such applications require monitoring the network performances to control content quality, thus guarantee quality of service (QoS) and quality of experience (QoE). In this paper, we propose a history-based TCP throughput prediction method incorporating communication quality features using SVR (Support Vector Regression). By taking history of communication quality features such as historical throughput and Received Signal Strength Indication (RSSI) into consideration, the throughput prediction error can be decreased. We conduct experiments with the proposed method and compare the prediction accuracy with a variety of methods in different scenarios of various moving modes of users. Results show that the proposed model could predict throughput effectively in various scenarios and decrease throughput prediction errors by a maximum of 26.47% compared with other methods.

The increasing demand of mobile applications has brought large amount of mobile traffic. To meet users' requirements for high-quality video delivery, it is an urgent task to provide fair-quality video delivery for various users and situations. In this paper, we evaluate QoS and QoE characteristics and validate QoE unfriendliness in heterogeneous DASH contents distributions to provide QoE-fair video delivery,. In the evaluations, we employ multiple contents with different resolutions, frame rates, and rate-distortion characteristics. By using heterogeneous DASH contents, we evaluate the effect of playout buffer size on QoS and QoE performances. Evaluation results show that smaller playout buffer size contributes to QoE improvement under network congestion. In addition, we confirm that suppression of playback stall is a particularly important factor to achieve QoE fairness compared to other QoS metrics such as initial delay and representation.

In recent years, various approaches have been investigated towards blind image quality assessment (IQA) with high accuracy and low complexity. In this paper we develop a pre-saliency map based blind IQA method, which takes advantage of saliency information in prior of quality prediction for performance enhancement by two steps. 1) We split the image into patches and design a convolution neural network (CNN) to predict the patch-wise quality score. Then we explore the relation between image saliency information and CNN prediction error to present a statistical analysis. 2) Based on the analysis, we propose a patch quality aggregation algorithm by removing non-salient patches which are likely to bring large prediction error and assigning large weights for salient patches. Experimental results validate that our method can achieve high accuracy (0.978) with subjective quality scores, which outperforms existing IQA methods. Meanwhile, the proposed method can reduce 52.7% computational time than the IQA without pre-saliency map.

In this paper, we performed human moving pattern recognition using communication quality: cellular download throughputs, Received Signal Strength Indicators (RSSIs) and cellular base station IDs. We apply three machine learning algorithms, such as Support Vector Machine (SVM), K-Nearest Neighbor (KNN) and Random Forest (RF) and evaluate recognition accuracy of human moving patterns. Results conclude that the communication quality can recognize moving patterns with high accuracy.

In this paper, to provide high-quality and high-visibility surveillance, we introduce video surveillance system equipped image enhancement system. Because quality of enhanced image heavily depends on encoding characteristic, we evaluate the subjective and objective image quality of image enhancement under various video encoding rates by using night and day sequences.

In this paper, we evaluate QoS (Quality of Service) and QoE (Quality of Experience) characteristics of 2K and 4K video distribution using MPEG-DASH (Dynamic Adaptive Streaming over HTTP). Evaluation results conclude that smaller playout buffer size contributes to QoE improvement.

Because mobile users demand high quality and energy-friendly video delivery service that efficiently uses wireless resources, we introduce a green video delivery system using moving route navigation and playout buffer control based on the mobile throughput history data. Our system first determines the optimal travel route to realize high-speed and energy-efficient communications. Then when a user enters a high throughput area, our system temporarily extends the video playout buffer size, and if allowed, the user remains in the high throughput area and aggressively downloads video segments until filling the extended buffer via a high-speed and energyefficient wireless connection. After leaving this area, the user consumes video segments from the extended buffer in order to keep smooth video playback without wireless communications. Computer simulations along with laboratory and field experiments on the system performance confirm that our system can achieve green mobile video delivery.

In this study, we propose a broadcasting and communication integrated platform that can retransmit 8K broadcasting contents from a TV receiving set to mobile devices. We then evaluate the QoS characteristics and the qualities of 8K contents in this platform. Evaluation results conclude that 8K broadcasting contents can be retransmitted to mobile devices smoothly and their qualities are nearly as high as those of assumed 8K broadcasting service in Japan.

Nowadays, researches on accident prevention using train-mounted cameras had been progressing. Our proposed method considers temporal continuity between frames by using motion vectors in addition to conventional thresholding on similarity values obtained by a human detection method using HOG features. Experiments show the effectiveness of our method as compared with a previous method using only HOG features.

In order to analyze system latency of edge computing for multimedia data processing, we propose a model of system latency of two different distributed processing scenarios: pipeline and parallel processing. We validate the model by using actual cloud services and human detection processing.

Urban surveillance systems are being applied in a rapid pace with mature but inefficient solutions. The inefficiency is revealed with two aspects, too concentrated bandwidth and processing requirement. To solve this problem, we proposed a content oriented surveillance system based on Information-Centric Network. However, we can't simply replace TCP/IP streaming structure with named contents streaming structure because it can't improve the surveillance system's efficiency enough. In this paper, we took the ICN network's profits even further with the named contents. Instead of streaming live video to the central data center and processing multiple data stream in the same time, we have designed the nodes to process the captured raw data and produce objective contents for the central data center. With the extremely size difference in raw data and actual valued contents from it, we could apply the method in investigating area people traffic conditions and even in disaster and anti-terrorism scenarios. There was a field experiment performed to evaluate tourists' densities and dressing habits during winter season of March. The experiment expressed the benefits of our system and compared our method with traditional surveillance systems in saving network bandwidth and functionalities.

In recent years, video sequence with a depth map can be taken by a camera with RGB-D (depth) sensor like Kinect. On the other hand, quality degradation at occluded regions is inevitable when conventional 2-D video coding is applied. In this paper, we propose usage of an occluded region detection supported by the depth map for improvement of 2-D image coding quality. Experiments show that further improvement of picture quality at the whole of frame and occluded regions is addressed for H.265/HEVC compressed streams.

Image super-resolution quality is degraded by the noise component of an image. However the noise should not be eliminated because of its importance in high-definition impression for digital cinema. We therefore propose an image super-resolution method using the synthesis of the super-resolution signal and noise components respectively after dividing an original image into signal and noise components. The signal component's resolution is enhanced using registration between the signal component and its multi-scale Wavelet components with resolution-enhanced parameter optimization.

Applications and malware affecting them are dramatically changing. It isn't certain whether the currently used features can classify normal traffic or malware traffic correctly. In this paper, we evaluated the features used in previous studies while taking into account secular changes to classify normal traffic into the normal category and anomalous traffic into the anomalous category correctly. A secular change in this study is a difference in a feature between the date the training data were caputred and the date the test data were captured in the same circumstance. The evaluation is based on the Euclidean distance between the normal codebook or anomalous codebook made by vector quantization and the test data. We report on what causes these secular changes and which features with little or no secular change are effective for malware detection. © 2013 Springer-Verlag Berlin Heidelberg.

Application-layer multicast has been considered as a promising method for streaming live media to users in P2P networks. However, the inefficient bandwidth utilization and the vulnerability to peer dynamics remain important issues. In this paper, we introduce a new approach to constructing an efficient overlay leveraging the strong features of tree and mesh methods. A core overlay is built in a fashion of multiple sub-trees, which allows each peer to receive data from other peers located in any different sub-trees. We propose a set of auxiliary connections, which connect between two nodes that do not have permanent parent/child relationship, to achieve a push-based data transmission from multiple parents without redundancy. Simulations through ns2 demonstrate that our scheme provides a good streaming quality with low end-to-end delay. Also, the proposed auxiliary links is beneficial in improving the resiliency of the overlay and utilizing the bandwidth of peers.

The use of mobile terminals has been expanding dramatically in recent years as they evolve from a means of dispatching and gathering information to a highly functional tool that supports personal lifestyles and behavior. A mobile terminal is likely to store various kinds of personal information such as a calendar and contact information as well as key data to carry out online transactions. Losing one's mobile terminal therefore creates the possibility that one's personal information may fall into the wrong hands and be used for malicious purposes.
We therefore propose a method of personal authentication using sensor data in a mobile terminal. First, we applied the LPC cepstrum to this authentication and checked for validity. We also evaluated the effectiveness of gait authentication using several frames.

We propose a novel image super-resolution method from digital cinema to 8K ultra high-definition television using registration of wavelet multi-scale components with affine transformation. The proposed method features that an original image is divided into signal and noise components by the wavelet soft-shrinkage with detection of white noise level. The signal component enhances resolution by registration between a signal component and its wavelet multi-scale components with affine transformation and parameters optimization. The affine transformation enhances super-resolution image quality because it increases registration candidates. The noise component enhances resolution with power control considering cinema noise representation. Super-resolution image outputs by synthesis of super-resolved signal and noise components. Experiments show that the proposed method has objectively better PSNR measurement and subjectively better appearance in comparison with conventional super-resolution methods. © 2013 IEEE.

In this paper, an image noise in the digital cinema is analyzed at first. Based on this analysis, we propose the image super-resolution method from digital cinema to ultrahigh definition television. The proposed method separates the digital cinema into signal and noise components by the spatio-temporal Wavelet analysis, super-resolves these components respectively and synthesize them.

In this paper, we propose a system which automatically generates slideshows for music, by utilizing images retrieved from photo sharing web sites, based on query words extracted from song lyrics. The proposed system consists of two major steps: (1) query extraction from song lyrics, (2) image selection from web image search results. Moreover, in order to improve the display duration of each image in the slideshow, we adjust image transition timing by analyzing the duration of each lyric line in the input song. We have conducted subjective evaluation experiments, which prove that the proposal can generate impressive music slideshows for any input song. © 2010 International Society for Music Information Retrieval.

本稿では量子化制約条件を考慮した複数フレームの合成によりPSNRを改善する手法を提案する。複数フレームの合成の際には同一領域を合成する必要があり、動き検出の性能が全体の性能の向上に繋がると考えられる。そこで動き検出において、いくつかの改善手法を示す。また、反復処理における問題点についても述べた後、改善手法として量子化の制約条件を外れた場合に行うDCT係数の補正処理を制限する手法について説明し、それぞれシミュレーション実験によりその有用性を示す。

This research presents a formal user evaluation of a typical visualization method for content-based music information retrieval (MIR) systems, and also proposes a novel interface to improve MIR usability. Numerous interfaces to visualize content-based MIR systems have been proposed, but reports on user evaluations of such proposed GUIs are scarce. This research aims to evaluate the effectiveness of a typical 2-D visualization method for content-based MIR systems, by conducting comparative user evaluations against the traditional list-based format to present MIR results to the user. Based on the observations of the experimental results, we next propose a 3-D visualization system, which features a function to specify sub-regions of the feature space based on genre classification results, and a function which allows users to select features that are assigned to the axes of the 3-D space. Evaluation of this GUI conclude that the functions of the 3-D system can significantly improve both the efficiency and usability of MIR systems. © 2009 International Society for Music Information Retrieval.

In this paper, we introduce a new mesh-based approach for data delivery which is organized over multiple tree-shaped core routes. Given that both tree and mesh approaches have their own strong points, we simply combine them together. We evaluate the proposal through ns-2 simulator. The simulation results demonstrate that our approach can provide higher average received quality and has acceptable data delivery delay when compared with a single tree method. We also show that, over a static overlay, the push-based data delivery on mesh can provide the received quality close to pull-based data delivery method with less latency. As well, it has lower control overhead than the pull-based method when the peer number is large.

Recently it has been shown that scalable (layered) image is appropriate for the Internet because of its better flexibility and functionality. At the same time, the pre-fetching technology also proves to be an efficient method to improve the performance of image-delivery with the low delay and adjustable quality. However, how to deploy the pre-fetching to deliver scalable contents has not been mentioned. Therefore, this paper presents an analysis of the pre-fetching algorithm to deliver layered image contents over the Internet. Based on the analysis of web request probability and the relations between different layers in the scalable imaging delivering system, a new pre-fetching algorithm is proposed to drastically improve cache performance. This proposal is a request frequency-based approach, in which a progressive image format can be used.

In order to deliver multimedia contents to the end-users with low delay and acceptable quality, the pre-fetching technologies have been shown the efficiency if we can predict user&apos;s future requests and prefetch the data needed previously. However, because both the network topology and Web access are changed dynamically, whether the pre-fetching can get the stable performance under dynamic environment becomes an important issue. Therefore, this paper talks about the robustness of the pre-fetching algorithm, which has been proposed by us to deliver layered image contents over the Internet. Based on the web access distribution, the request pattern in the scalable imaging delivering system is analyzed in this paper. Furthermore, we check the robustness of our proposal by changing related parameters. Simulation results show that Our proposal is stable against variable conditions even if access patterns are dynamically changed.

The total amount of information to process or analyze is jumping sharply with the quick spread of computers and networks. Our project, «Highly scalable monitoring architecture for information explosion», develops a monitoring system allows observing systems, merging the system logs, and discovering intelligence to share. More concretely, the project builds the total system to maintain, optimize, and protect autonomically. This paper reports the outcomes of the project after first-half of the development period.The rest of the paper is organized as follows. Section 2 describes the concept and details of the monitoring system on a single node, and Section 3 addresses the aggregation of the collected information in distributed environments. Section 4 and Section 5 introduce applications of the monitoring systems. Section 6 summarizes the project and mentions future plans. © 2009 IEEE.

In this paper, we introduce a new mesh-based approach for data delivery which is organized over multiple tree-shaped core routes. Given that both tree and mesh approaches for multicast overlay construction have their own strong points, we simply combine them. During session, the mesh connections are formed dynamically to facilitate the data delivery on tree-shaped core routes which may experience bandwidth bottlenecks. Unlike existing mesh-based approaches, our proposal employs push-based data delivery along mesh connection. A well-organized mesh construction scheme is defined to avoid data redundancies. We evaluated the proposal through ns-2 simulator. The simulation results demonstrate that our approach can provide higher average received quality and has acceptable control overhead when compared to a single tree method. We also show that, over a static overlay, the push-based data delivery on mesh can provide the received quality close to supposed pull-based data delivery method with less latency. Moreover, it has lower control overhead than the pull-based method when the peer number is large.

Content Delivery Overlays improves end-user performance by replicating Web contents on a group of geographically distributed sites interconnected over the Internet. However, with the development whereby overlay systems can manage dynamically changing files, an important issue to be resolved is consistency management, which means the cached replicas on different sites must be updated if the originals change. In this paper, based on the analytical formulation of object freshness, web access distribution and network topology, we derive a novel algorithm as follows: (1) For a given content which has been changed on its original server, only a limited number of its replicas instead of all replicas are updated. (2) After a replica has been selected for update, the latest version will be sent from an algorithm-decided site instead of from its original server. Simulation results verify that the proposed algorithm provides much better consistency management than conventional methods with the reduced the old hit ratio and network traffic.

The purpose of this study is to maintain efficient backup routes for reconstructing overlay trees quickly. In most conventional methods, after a node leaves the trees, its child nodes start searching for the new parents. In this reactive approach, it takes a lot of time to find a new parent. In this paper, we propose a proactive approach to finding a next parent as the backup route node over the overlay tree before the current parent leaves. A proactive approach allows a node to find its new parent node immediately and switch to the backup route node smoothly. In our proposal, the structure of the overlay tree using a redundant degree can decide a backup route node without so much overhead. Simulations demonstrate our proactive approach can recover from node departures 2 times faster than reactive approaches, and can construct overlay trees with lower overheads than another proactive method. Additionally we carried out experiments over actual networks and their results support the effectiveness of our approach. We confirmed that our proposal achieved better streaming quality than conventional approaches.

This paper proposes a feature space modification method for feature extraction of music, which is effective for the development of a content-based music information retrieval (MIR) system based on user preferences. The proposed method conducts clustering of all songs in the music collection, and utilizes the resulting cluster IDs as training data for feature space modification. and is capable to automatically generate a feature space which is suitable to the content of any music collection. Experiment results prove that the proposed method improves accuracy of user preference based MIR.

Content Distribution Overlays improves end-user performance by replicating Web contents on a group of geographically distributed sites interconnected over the Internet. However, with the development whereby overlay systems can manage dynamically [2] changing files, an important issue to be resolved is consistency management, which means the cached replicas on different sites must be updated if the originals change. In this paper, based on the analytical formulation of object freshness time, web access distribution and network topology, we derive a novel algorithm as follows: (1) For a given content which has been changed at its original server, only a limited number of its replicas instead of all replicas are updated. (2) After a replica has been selected for update, the latest version will be sent from an algorithm-decided site instead of from its original server. Simulation results verify that the proposed algorithm provides much better consistency management than conventional methods with the reduced update overhead and network traffic.

This paper proposes a feature space modification method for feature extraction of music, which is effective for the development of a content-based music information retrieval (MIR) system based on user preferences. The proposed method conducts clustering of all songs in the music collection, and utilizes the resulting cluster IDs as training data for feature space modification, and is capable to automatically generate a feature space which is suitable to the content of any music collection. Experiment results prove that the proposed method improves accuracy of user preference based MIR.

The purpose of this study is to maintain efficient backup routes for restoring overlay trees. In most conventional methods, after a node leaves the trees, its children start searching for a new parent. In this reactive approach, it takes a lot of time to find a new parent. In this paper, we propose a proactive approach to find a new parent over the overlay trees before the current parent leaves. A proactive approach can find respective new parents immediately and switch to the backup route smoothly. In our proposal, the structure of the overlay tree using a redundant degree enables to decide a new parent without so much overhead information. Simulations demonstrate our proactive approach can recover from node departures 2 times faster than reactive approaches, and can construct overlay trees with lower overheads than another proactive method. Additionally we carried out experiments over actual networks and their results support the effectiveness of our approach. We confirmed that our proposal achieved better streaming quality than conventional approaches.

This paper proposes a novel on-demand multipath routing protocol for a mobile ad hoe network. By applying a newly developed route update procedure with combined metrics of delay, hop count and disjointness, each intermediate node deliberately selects multipath candidates while contributing to suppression of unnecessary routing packets. Extension of RREQ/RREP packets with a source route list is also incorporated, not only to alleviate limitation of the hop-count based approaches but rather to provide more efficient multiple routes. Computer simulations using ns-2 simulator are performed with comparison to conventional methods and effectiveness of the proposed methods is quantitatively validated.

CDN (Content Delivery Networks) improves end-user performance by replicating web contents on a group of geographically distributed servers. However, repeatedly keeping the entire replica of the original objects into many content servers consumes too much server resource. This problem becomes more serious for the large-sized objects such as streaming media, e.g. high quality video. In this paper, we therefore propose an efficient replication method for layered video streams in CDN, which can reduce user response delays and storage costs simultaneously. Based on an analytical formulation of the cooperative replication of layers and segments of each video stream, we derive a replication algorithm which solves next three problems quantitatively. (1) How many servers should be selected to replicate a given video stream? (2) For a single video stream, how many layers and segments should be stored in a given server? (3) After selecting a group of servers for each video stream, how do we allocate the replication priority (i.e. order) to each server? Simulation results verify that the proposed algorithm efficiently resolves the above problems and provides much better performance than conventional methods.

CDN (Content Delivery Networks) improves end-user performance by replicating web contents on a group of geographically distributed servers. However, repeatedly keeping the entire replica of the original objects into many content servers consumes too much server resource. This problem becomes more serious for the large-sized objects such as streaming media, e.g. high quality video. In this paper, we therefore propose an efficient replication method for layered video streams in CDN, which can reduce user response delays and storage costs simultaneously. Based on an analytical formulation of the cooperative replication of layers and segments of each video stream, we derive a replication algorithm which solves next three problems quantitatively. (1) How many servers should be selected to replicate a given video stream? (2) For a single video stream, how many layers and segments should be stored in a given server? (3) After selecting a group of servers for each video stream, how do we allocate the replication priority (i.e. order) to each server? Simulation results verify that the proposed algorithm efficiently resolves the above problems and provides much better performance than conventional methods.

TCP is a popular internet protocol for reliable end-to-end data delivery, but it cannot be directly applied to wireless networks in which packet loss may be induced by higher BER or handover than congestion. TCP assumes that such packet loss is caused by network congestion and initiates congestion control procedures. In this paper, we present a novel protocol using Explicit Handover Notification to improve TCP performance over wireless links. Additionally, we execute computer simulations using network simulator and compare with other various protocols.

This paper presents a method of gender classification by integrating facial, hairstyle, and clothing images. Initially, input images are separated into facial, hairstyle and clothing regions, and independently learned PCAs and GMMs based on thousands of sample images are applied to each region. The classification results are then integrated into a single score using some known priors based on the Bayes rule. Experimental results showed that our integration strategy significantly reduced error rate in gender classification compared with the conventional facial only approach.

In this paper, we propose a motion vector search algorithm that employs the Sum of Absolute Difference (SAD) minimization in DCT domain as a distortion measure. In addition, block based motion estimation is carried out with multiple modes in this frequency domain, and a mode is selected according to the Lagrangian cost function [1, 2, 3]. Four modes are designed. The first two modes are a mode with 1 motion vector (INTER1V) and a mode with 4 motion vectors (INTER4V) per macroblock that are similar to the H.263 [4]. The last two modes use 4 motion vectors per macroblock (INTER1V+) or per block (INTER4V+), where the 8×8 DCT coefficients are firstly separated into 4 groups and a motion vector is independently assigned to each group. The average SNR (Signal to Noise Ratio) gains between 0.3 and 1.0dB are obtained for tested sequences when compared with TMN. Our proposal method provides true motion vectors for low frequency and minimizes distortion for higher frequency. Therefore our proposed method also reduces subjective visual degradation.

CDN (Content Delivery Networks) improves end-user performance by replicating web contents on a group of geographically distributed content servers. Replication Algorithm plays an important role in helping users to retrieve Web objects from the content servers. If a user can directly get the requested objects from the content server, he need not to contact the remote origin server and the user delay can be reduced. However, current replica strategies in CDN are to simply and repeatedly keep the complete replica of the original object on many content servers. This method has some disadvantages, including too much consumed server space and a waste of the storage cost. It is more serious for replicating some large-sized objects such as streaming media, which are being distributed over the Internet more and more. In this paper, we discuss a replication strategy for scalable video streaming in CDN to reduce user response and storage cost. Based on theoretical analysis, assuming layered video coding, we propose a novel replication algorithm which deals with following three problems. (1) How many content servers should be selected to replicate a given video content? (2) For a single video content, how many layers should be kept in a given content server? (3) After selecting a group of content servers for each video content, how do we decide the replication priority for each content server? Simulation results show that the proposed algorithm can efficiently resolve the above problems, and provide much better performance than the conventional methods.

With the advance of high-speed network technologies, availability and popularity of streaming media contents over the Internet has grown rapidly in recent years. Because of their distinct statistical properties and user viewing patterns, traditional delivery and caching schemes for normal web objects such as HTML files or images can not be efficiently applied to streaming media such as audio and video. In this paper, we therefore propose an integrated caching scheme for streaming media with segment-based caching and hierarchically distributed proxies. Firstly, each stream is divided into segments and their caching algorithms are considered to determine how to distribute the segments into different level proxies efficiently. Secondly, by introducing two kinds of segment priorities, segment replacing algorithms are proposed to determine which stream and which segments should be replaced when the cache is full. Finally, a Web-friendly caching scheme is proposed to integrate the streaming caching with the conventional caching of normal web objects. Performance of the proposed algorithms is verified by carrying out simulations.

The efficient distribution of stored information has become a major concern in the Internet. Since the web workload characteristics show that more than 60% of network traffic is caused by image documents, how to efficiently distribute image documents from servers to end clients is an important issue. Proxy cache is an efficient solution to reduce network traffic. And it has been shown that an image caching method (Graceful Caching) based on hierarchical coding format performs better than conventional caching schemes in recent years. However, as the capacity of the cache is limited, how to efficiently allocate the cache memory to achieve a minimum expected delay time is still a problem to be resolved. This paper presents an integrated caching algorithm to deal with the above problem for image databases, web browsers, proxies and other similar applications in the Internet. By analyzing the web request distribution of the Graceful Caching, both replacing and pre-fetching algorithms are proposed. We also show that our proposal can be carried out based on information readily available in the proxy server; it flexibly adapts its parameters to the hit rates and access pattern of users&#039; requesting documents in the Graceful Caching. Finally we verify the performance of this algorithm by simulations.

The emergence of the Internet as a pervasive communication medium has led to the rise of many applications of streaming media. However, because of their distinct statistical properties and user viewing patterns, traditional delivery and caching schemes for web objects such as HTML files or images can not be efficiently applied to streaming media such as audio and video. In this paper, we therefore propose an integrated caching scheme for streaming media with segment-based caching and with hierarchically distributed proxies. Firstly, each stream is divided into segments and their caching algorithms are considered to determine how to distribute the segments into different level proxies efficiently. Then, by introducing two kinds of segment priorities, segments replacing algorithms are proposed to determine which stream and which segments should be replaced when the cache is full. Finally, a web friendly caching scheme is proposed to integrate the streaming caching with the conventional caching of normal web objects. Performance of the proposed algorithms is verified by carrying out simulations.

The efficient distribution of stored information has become a major concern in the Internet. Since the web workload characteristics shows that more than 60% of network traffic is caused by image, documents, how to efficiently distribute image documents from servers to end clients is an important issue. Proxy cache is an efficient solution to reduce network traffic. An it has been shown that an image caching method (Graceful Caching) based on hierarchical coding format showed better performance than conventional. caching schemes in recent years. However, as the capacity of the cache is limited, how to efficiently allocate the cache memory to achieve a minimum expected delay time is still a problem to be resolved.
This paper presents an integrated caching algorithm to deal with the above problem in the Internet. By analyzing the web request distribution of Graceful Caching, both replacement and pre-fetching algorithms are proposed. We also show that our proposal can be carried out based on information readily available in the proxy server and it flexibly adapts its parameters to the hit rates and access pattern of users' requesting documents in the Graceful Caching. We finally verify the performance of this algorithm by simulations.

This paper presents a cache scheme for image databases, web browsers, proxies and other similar applications in Internet The current cache system employs a hard strategy: either the image Is stored in a cache or not even its data is quite big. We have proposed a caching scheme featuring human memory mechanism (image database system featuring graceful oblivion), which is called Graceful Caching. The previous work shows that Graceful Caching reduces user response time by assigning a variable amount of memory to each image. Several recent studies suggest that prefetching, techniques could be employed to further improve the cache performance, by anticipating and prefetching future client requests. Some approaches based on request frequency have been proposed to the conventional caching (hard system) and got better results. In this paper, we try to apply a request frequency based approach. to Graceful Caching system, in which a progressive image format is used. Based on the analysis of web request probability in Graceful Caching, we go on to propose a new prefetching algorithm that drastically improves cache performance of system. We verify the performance of this algorithm by simulations.

A new wavelet coding method is presented to exploit both intra/inter-band correlations. While the traditional zerotree approach can exploit correlation between subbands, a wavelet decomposition has properties more suitable for the runlength coding method, which can exploit spatial correlation. The proposed method classifies wavelet coefficients into groups according to their parents' magnitude. Each group is chosen in order of the magnitude and scanned to keep spatial correlation. It outperforms the zerotree-based coding approach especially at low bit rate.

This paper describes novel methods to improve coding efficiency of the wavelet transform from the viewpoint of scanning techniques. Firstly, statistical properties of the wavelet transform are examined using real image sequences. They indicate energy distribution or correlations of wavelet transformed coefficients. Taking them into account, efficient scanning methods are investigated for two different approaches. The first approach treats transform coefficients, located in spatially neighboring position, in 8 × 8 blocks. This is called block based approach, and its optimum scanning path is examined similar to the zigzag scan for DCT (discrete cosine transform). Furthermore, a new scanning algorithm utilizing correlation between transform coefficients is proposed. The second approach carries out scanning over the entire subimage created by the wavelet transform. This is called subband based approach, in which scanning is applied to each subband. A new scanning algorithm is presented utilizing a block-wise non-linear filter which removes isolated non-zero coefficients. Simulations are carried out, and their results show the effectiveness of the above approaches especially for low bit-rate applications.

This paper presents novel algorithms exploiting multiple camera inputs and segmentation techniques, which can be applied to image fusion, disparity detection and object extraction. Differently focused images, stereo pairs and both of them are used for fusion, disparity detection and object extraction, respectively. Firstly, image fusion is done by segmentation of each image and determination of focused regions per segment. An efficient decision criterion is developed taking a method of auto-focus into consideration. Secondly, disparity detection is executed by recursively applying segmentation and disparity detection per segment. A new clustering criterion is proposed in order to achieve good segmentation and high compression ratio of disparity maps simultaneously. Finally, object extraction is carried out by utilizing both the fusion result and the disparity map. Experiments are carried out, and they show us effectiveness of the proposed algorithms.

This paper presents a novel framework for structure recovery utilizing a scaled orthographic view and perspective views simultaneously. Perspective views lead to precise recovery based on the triangulation principle; however, many parameters such as camera positions, camera poses and focal length, must be measured beforehand. Thus, a scaled orthographic view is introduced as a subsidiary system to achieve them automatically. Camera parameters ate calculated implicitly owing to the scaled orthographic properties, and then the structure recovery is done by the orthogonality of camera coordinate systems and the triangulation principle. The experimental results provide sufficient accuracy of structure recovery.

This paper presents a novel framework which proves the superiority of overlapped motion compensation. Window design problem is revised by introducing a statistical model of motion estimation process. The result clarifies relationship between the optimum window and image characteristics in an explicit formula and quantifies prediction error reduction achieved by overlapped motion compensation. Experimental results using real image sequences support the proposed theory and demonstrate its superiority. Overlapping in warping prediction is also considered and its effectiveness is shown.

This paper presents mathematical frameworks on temporal predictive processing in the MPEG video compression standard. Firstly, a coding gain is derived based on traditional prediction theories. The optimum ordering of three different picture types (I,P,B-pictures) is clarified according to image source characteristics. Secondly, a novel framework on the target bit assignment is presented with some experimental backgrounds. The solution consists of simple formulae, but brings drastic SNR gains to the conventional TM5 algorithm.

This paper describes theories on overlapped motion compensation and their applications to a wavelet codec aimed for very low bit-rate environment below 64 kb/s. The theories are concerned with the evaluation of prediction efficiency improved by overlapped motion compensation and also its smoothing effect on the discontinuities at block boundaries encountered when motion vectors do not coincide among neighboring blocks. They contribute to determine the optimum window shape for overlapped motion compensation in a developed wavelet codec, which suffers from reduction of coding efficiency when there are such discontinuities in a signal to be transformed. Regarding the wavelet codec, a new scanning method of transform coefficients and alternate use of normal and spatially-reversed basis functions for cascaded wavelet transform are introduced. Finally, the implementation of the codec on a Video Image Signal Processor (VISP) is discussed. © 1994 IEEE

This paper presents a filter design algorithm in order to realize high efficient image codec system based on subband decomposition, and also describes unified frameworks to interpret various image coding techniques. First, various image coding techniques based on the linear transform are dealt with in an unified manner, where both the matrix representation in time domain and the multirate filter bank concept are introduced. Reconsideration about the DPCM is stressed here because the subband coding and the transform coding have been already formulated in some common frameworks. Neither the filter&#039;s frequency response nor its orthogonality are taken into consideration because they don&#039;t necessarily play an important role in our formulation, particularly in case of the DPCM and the SSKF. Secondly, statistical optimization of short tap and linear phase PR (perfect reconstruction) filter banks is considered. The UCG (unified coding gain) that we proposed at the VSIP &#039;91 as a performance measure of energy compaction properties of the multirate filter bank is expanded to multiple layer cases, and several new examples are presented. The short tap structure leads to low computational complexity, and the linear phase property contributes not only to remove phase distortion but also to solve the so-called border problem. Simulation results are also shown, and the validity of our approach is confirmed.

In this paper, two analytical methods for evaluating the coding efficiency of subband coding are proposed, and optimization of filter coefficients of the perfect reconstruction FIR filter banks is considered, based on a new performance measure called unified coding gain. First, matrix representation of the subband coding in the time domain is considered, and conventional subband filter banks are classified into orthogonal ones such as the QMF and nonorthogonal ones such as the SSKF. For the orthogonal filter banks, the coding gain shown by Jayant and Noll is introduced, and their theoretical performance evaluation is carried out. However, this first method cannot be applied to nonorthogonal filter banks any longer because the coding gain is defined on the assumption of filter orthogonality. Therefore, an optimum bit allocation problem for subband coding is considered, and the unified coding gain, which can be applied to arbitrary subband filter banks, is derived as a new performance measure to take the place of the coding gain. This second method enables us to estimate the coding efficiency of arbitrary transform techniques as well as the subband approaches, and its result suggests that the SSKF(5 × 3) outperforms the QMF as long as the number of subbands is not too large, even though its filter length is much shorter. This result encourages us to find filter coefficients that maximize the unified coding gain according to filter length. In addition, new perfect reconstruction FIR filter banks which have not only low computational complexity but also good energy compaction properties are presented. © 1991.

One of the merits with the introduction of variable bit-rate coding to moving pictures is the constancy of picture quality. However, considering the fact that conventional SNR cannot reflect the result of subjective assessment sufficiently, other quality measures should be considered to design the variable bit-rate coders. In this paper, weighted SNR (WSNR), which is calculated by weighting coding errors with the spatial frequency response of human vision, is introduced as a quality measure in place of SNR, and the variable bit-rate coder based on this measure is considered. Then, subjective assessment is carried out and the effectiveness of our approach is confirmed. © 1991.

In this paper, two analytical methods to evaluate coding performance of subband coding are proposed, and optimization of its filter coefficients from the viewpoint of energy compaction property is considered. The first method is based on matrix representation of subband coding in time domain, where the coding gain given by Jayant and Noll is introduced as a performance measure for filter banks with orthogonal property. The second method is based on the optimum bit allocation problem for subband coding (multirate filter bank), where the unified coding gain is derived as a new performance measure which can be applied to arbitrary transform techniques. We then try to find filter coefficients which maximize the unified coding gain according to input characteristics. This approach leads to optimization of filter coefficients from the viewpoint of energy compaction property.

As a promising means of realizing broadband ISDN, the ATM (asynchronous transfer mode), which is based on fast packet switching, is considered interesting. In video communication by ATM, the constancy and improvement of picture quality by variable rate coding is expected, while the quality deterioration by cell loss is a serious problem. This paper proposes a video coding method introducing hierarchical coding to cope with the cell loss problem. The proposed hierarchical coding extracts the low‐frequency components of the image by iterating the hierarchical processing based on the spatial filter. Then it is possible to suppress the effect of the quality deterioration caused by cell loss, as long as the low‐frequency components are retained. First, the configuration of the proposed system is discussed. Simulations are executed to observe the effect of the cell loss and the coding characteristics, demonstrating the effectiveness of the proposed method in coping with the cell loss. Then discussions are made on the optimal number of hierarchical processings to cope with the cell loss and the protection level for hierarchical information. Furthermore, it is shown that a new type of video communication can be realized by introducing hierarchical coding. Copyright © 1990 Wiley Periodicals, Inc., A Wiley Company

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.

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.

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.

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.

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.

Image quality assessment (IQA) is an inherent research topic in image processing field for several decades. Recently, machine learning has achieved success in many multimedia tasks and can be applied in IQA. Especially, screen content images (SCIs) is greatly increasing in various applications, but the characteristics of SCIs makes it difficult to directly apply general IQA methods to predict qualities. In this paper, we propose a fast no-reference SCIs quality prediction method. First, we use the convolutional neural networks (CNNs) to predict the quality scores of each patch. Second, we present a SCIs-oriented quality aggregation algorithm for acceleration. Experimental results demonstrate that our method can achieve the high accuracy (0.957) with subjective quality scores, outperforming existing methods. Moreover, our method is computationally appealing, achieving flexible complexity performance by selecting different groups of patches.

The popularity of high dynamic range (HDR) makes the inverse tone mapping become an important technique for HDR display. In this paper, we propose a convolutional neural network (CNN) based inverse tone mapping method to generate a high-quality HDR image from one single standard dynamic range (SDR) image. First, we present a CNN design with a three-channel input, which considers both luminance and chrominance. Second, we propose to use overlapped inputs to remove the boundary artifacts, caused by zero padding in CNN. Experimental results demonstrate the high quality of our generated HDR images compared to the ground truth and conventional inverse tone mapping methods.

In order to recognize the transportation modes without any additional sensor devices, we propose a recognition method by using communication quality factors. In the proposed method, instead of Global Positioning System (GPS) and accelerometer sensors, we collect mobile TCP throughputs, Received Signal Strength Indicators (RSSIs), and cellular base station IDs (Cell IDs) through in-line network measurement when the user enjoys mobile services, such as video streaming service. In accuracy evaluations, we conduct two different field experiments to collect the data in five typical transportation modes (static, walking, riding a bicycle, a bus and a train,) and then construct the classifiers by applying Support Vector Machine (SVM), k-Nearest Neighbor (k-NN) and Random Forest (RF). Results conclude that these transportation modes can be recognized by using communication quality factors with high accuracy as well as the use of accelerometer sensors.

In this paper, to sustain a high-quality 24-hour video surveillance (i.e., high reliability) and reduce redundant video traffic volume (i.e., network friendliness), we propose an edge-centric video surveillance system that provides flexible adaptive control of the image enhancement process and video quality based on an event-driven adaptation. In the proposed system, the video bitrate is adaptively controlled according to the contrast of captured videos and conditions in a monitored area (e.g., 'normal', 'caution', and 'alert'). To confirm the system performance, we evaluate objective image quality, accuracy of human detection and video traffic volume generated by the proposed system. Evaluations conclude that the system can reduce the video traffic while sustaining high-quality visibility.

Adaptive rate video streaming on a novel Internet architecture, called Named Data Networking (NDN), appears to be a promising technology. However, given the source of content on NDN is unknown to the consumer, rate adaptation becomes inefficient. Moreover, the traffic at some nodes holding popular contents still can be congested. In this paper, we utilize the Internet of Things (IoT) device for assisting content delivery at an NDN router that has low bandwidth. The location of IoT devices installation is determined by a metric calculated from link bandwidth, delay, and utilization. We performed the experiments using DASH-NDN-JS on the setup network. The results demonstrate that our proposal could support NDN router to deliver video content to additional consumers efficiently. Thus, the network scalability can be improved with a reasonable cost.

Named Data Networking (NDN) has been recently introduced as a new future network architecture. The end-to-end throughput estimation for adaptive bit-rate video streaming on NDN is one of the most challenging topics. Specifically, the end-to-end throughput estimation on NDN appears to be unreliable, since the provider of content is unknown to the consumer. Moreover, partial caching on NDN router's Content Store could temporarily lead to packet loss due to throughput overestimation. In this paper, we present an active Interest adaptation scheme which operates by proactively estimating the throughput in a hop-by-hop fashion. The consumer node is then assisted with the most recent available end-to-end bandwidth. Therefore, the video player can promptly adapt to the change of network condition. The implementation evaluation using NDN-JS and DASH-JS on the setup network demonstrate that our proposed solution provides better average stream bit-rate and consumes less network bandwidth than the traditional system.

Throughput prediction is essential for ensuring high quality of service for video streaming transmissions. However, current methods are incapable of accurately predicting throughput in mobile networks, especially for moving user scenarios. Therefore, we propose a TCP throughput prediction method TRUST using machine learning for mobile networks. TRUST has two stages: user movement pattern identification and throughput prediction. In the prediction stage, the long short-term memory (LSTM) model is employed for TCP throughput prediction. TRUST takes all the communication quality factors, sensor data and scenario information into consideration. Field experiments are conducted to evaluate TRUST in various scenarios. The results indicate that TRUST can predict future throughput with higher accuracy than the conventional methods, which decreases the throughput prediction error by maximum 44% under the moving bus scenario.

Throughput prediction contributes a lot to adaptive bitrate control, adjusting the quality of video streaming accordingly to offer smooth media transmission and save energy at the same time. To solve the problem of throughput prediction for real time communication, this paper puts forward a new history-based throughput prediction method applying Hidden Markov Model in mobile networks. The main purpose of this method is to predict future throughput for real time communication in mobile network. Our novel approach utilizes Hidden Markov Model (HMM) with Gaussian Mixture Model (GMM) to deal with history time series of throughput and judge fluctuation factor with total variance when predicting future throughput. By conducting experiments with the new methodology, we compare the accuracy of the proposed method with three other conventional prediction methods. Results show our proposed method could identify data fluctuation effectively and predict future 100s throughput with high accuracy in various situations.

With increase of mobile wireless LAN systems, frequency contamination by the overlapping basic service set (OBSS) becomes a critical issue. In this paper, we focus on HCF controlled channel access (HCCA) to alleviate the OBSS problem. HCCA considers single BSS (SBSS) environment only and suppresses traffic of co-existing DCF (distributed coordination function) based WiFi access points (APs). We propose two methods that utilize capture effects to improve the coexistence capability with DCF networks, that we call DCF friendliness. The first method adjusts transmission timing in HCCA WLAN by applying inter-AP coordination. The second method enables simultaneous communication with surrounding DCF WLANs by changing the frame interval of HCCA to DIFS+1 SlotTime. Simulations show that both the proposed method can achieve higher throughput and better DCF friendliness.

In this paper, we propose a method to improve H.265/HEVC encoding performance for 8K UHDTV moving pictures by detecting amount or complexity of object motions. In 8K video, which has high spatial resolution, motion has a big influence on encoding efficiency. The proposed method estimates motion complexity by external process, and selects an optimal prediction mode and search ranges of motion vectors. By this method, we achieved highly efficient and low computation encoding.

In order to provide high-quality and highly reliable video delivery services for mobile users, especially train passengers, we propose a proactive content caching scheme that uses transportation systems. In our system, we place content servers with cache capability [e.g., content centric networking/named data networking (CCN/NDN)] in every train and station. Video segments encapsulated by MPEG-Dynamic Adaptive Streaming over HTTP (MPEG-DASH) are distributed and pre-cached by the station servers before the trains arrive at the stations. The trains receive content via high-speed wireless transport, such as wireless LANs or millimeter waves, when they stop at the stations. We developed prototype systems based on hypertext transfer protocol and CCN/NDN protocol, evaluate their performance through two field experiments that uses actual trains, and compare with traditional video streaming over cellular networks. Such evaluations indicate that our system can achieve high-quality video delivery without interruption for up to 50 users simultaneously.

Information Centric Networking (ICN) is a new networking paradigm in which the network provides users with named content, instead of communication channels between hosts. However, many issues, such as naming, routing, resource control, and security, still need to be resolved before it can be realized practically. Further, the energy efficiency of ICNs has not been sufficiently considered. In this study, we evaluate the energy consumption of an ICN network by first measuring the power consumption of actual PCs/smartphones and constructing an energy consumption model. Then, by using the model and the simulator ndnSIM, we calculate the energy consumed by the network. We also evaluate the energy consumed for video delivery, with a view toward power-saving video delivery.

In order to tackle recent mobile traffic explosion, content offloading using Wifi access points from cellular network is deployed but the Wifi system itself is sometimes heavily congested. In this paper, we propose a content offloading system using transportation systems such as a train. In addition, we aim to incorporate CCN (Content Centric Networking) into our system to exploit its in-network caching capability. We compare three protocol designs, pure IP, IP+CCN hybrid and pure CCN, and show the efficiency of our system using a network simulator.

In recent years, researches of underwater sensor networks have continued to investigate environment and resources of the sea. Acoustic waves are used instead of radio waves for wireless communication in underwater. However, dedicated hardware is very expensive, experiments on the sea are very time-consuming, and huge water spaces are necessary to study the underwater acoustic communication. In this paper, we present a cheap and tractable software-defined acoustic communication platform running on PCs using MATLAB, and evaluate its characteristics in a variety of communication methods by changing modulation schemes, error correction codes, transmission power and frequency by using commercial speaker and microphone devices. Our current implementation achieves data rate of up to 4.5 Kbps.

A system is proposed that uses content-centric networking to deliver content to mobile users. It uses sensor data obtained from the device they are using to identify their means of transport. The content is then delivered in a manner that best matches that means. The means of transport is identified using a combination of features: average magnetic quantity, average value and standard deviation of acceleration, vehicle speed, and maximum power spectrum of acceleration.

An accurate (reliable) motion estimation (ME) algorithm using block matching for sequences including spatially periodic patterns is proposed and evaluated in comparison to conventional algorithms. Mismatch between accurate motion vectors (MVs) and MVs obtained by block matching (BM) which minimizes BM distortion is evaluated. BM distortion caused by MV sampling error (DVSE), which causes ME errors, is discovered. A new ME algorithm which uses the maximum value of DVSE (maxDVSE) is proposed. As for the proposed algorithm, maxDVSE is calculated for each block as a threshold of accurate vector candidates, and all candidate vectors for each block in a frame are stored. Accurate vectors are selected from the candidate vectors by finding so-called "anchor vectors" (namely, vectors at a unique point on spatially periodic patterns). The MVs are selected in a process starting from one block and propagated to neighboring blocks. Simulations are performed and the results show that average precision of the proposed ME algorithm is 98.9%, which is superior to that of other conventional algorithms.

Proactive content caching scheme utilizing transportation systems, especially on trains, was proposed in order to provide a robust content delivery with efficient wireless resource usage. This system requires content servers with NDN capability to be placed on every station and trains. The mechanism is to pre-cache the contents that users request to the station server before the train arrives, and the train server caches the content during its stoppage time. With this mechanism, users are able to have a continuous playback of videos while riding on trains. In this paper, we have proposed a browser-based implementation, called DASH-NDN-JS, for this proactive content caching scheme. We evaluate this scheme, and experiment with multiple users to see how it will affect the video quality each user will achieve and the bandwidth consumption between the connections. Our evaluations conclude that the increase of users lowers the video quality, but avoids congestion depending on what video content each user will want to request.

Block-based Motion Estimation (ME) has been widely used in various video coding standards to remove temporal redundancy. However, this ME has limitation that it can only compensate for a parallel translation. Various methods have been proposed for other motions such as zooming. In recent years, 8K UHDTV (7,680 x 4,320 pixels) has been developed. Since 8K has large motion by zooming that is difficult to be predicted by block matching, it is important to improve zoom motion estimation. In this paper, to handle zooming in 8K video sequences, we propose a method for improving the picture quality by global zoom estimation based on motion vector analysis extracted by block matching.

Providing robust content delivery service, such as Video on Demand, with efficient wireless resource usage is important for mobile users. To achieve this, we have proposed a proactive content caching scheme utilizing transportation systems (e.g., train). Because our previous prototype is immature and has a scalability issue, in this paper, we introduce an efficient transmission mechanism of Interest messages to improve throughput efficiency in mobile NDN, develop two prototypes based on NDN and HTTP, and then evaluate prototype performances by performing larger-scale field experiments on actual commercial railroad line. Evaluations conclude that our system can achieve high efficient content transfer and fully utilize wireless network bandwidth. As a result, our system can provide high-quality video streaming for up to 50 users simultaneously.

© Springer International Publishing Switzerland 2015. This paper presents a live version music identification system by modifying the conventional cover song identification system. The proposed system includes two stages: a live version identification phase and an audio scenedetection phase. We improve the accuracy of the system by weighting similarity scores in the live version identification phase and discriminating scenes by using RMS, pulse clarity and similarity scores. Results show that the proposed method performs better than the previous method. The final algorithm achieves 70% accuracy on average.

Providing robust video streaming along with efficient wireless resource usage is necessary for mobile users, especially on subway, and mobile carriers. To achieve this, we propose a highly-reliable buffer strategy based on long-term throughput prediction. Our approach has two elements which are called "long-term throughput prediction" and "guaranteed playout buffer filling mechanism." To avoid any video freeze due to network quality degradation, our approach calculates the optimal amount of playout buffer and schedules video download timing in a theoretical manner. We evaluate its performance via experiments in real environment. Evaluations conclude that our approach can provide highly-reliable video streaming and also achieve to reduce the average playout buffer size on the client.

In wireless networks, it is important to realize energy-efficient video delivery. To do this, we introduce energy-efficient video streaming over named data networking (NDN). In our proposed approach, we focus on two areas, namely, to improve the throughput performance by Interest aggregation, and to reduce the overhead energy using playout buffer-size control. We evaluate the power savings realized by our method using a hardware power-measurement tool. Our results show that Interest aggregation can realize an approximately 50% reduction in energy compared to conventional NDN implementation, and a large playout buffer size can reduce the energy by approximately 37% compared to the case of a small playout buffer size.

Content Centric Networking (CCN) has been emerged as a new networking paradigm. In CCN, adaptive video streaming is one of the most promising technologies. However, rate adaptation and outbound face selection are the primary concern since a node cannot reliably estimate the end-to-end throughput. In this paper, we propose a strategy for outbound face selection considering both response time and bit-rate of lost packets. These two aspects facilitate the node to choose which outbound face offers the best QoE. In a case where bit-rate of lost packets is greater than a half of transmitted bit-rate, the node will re-select outbound faces to avoid intermittent streaming. The experiments performed using CCNx demonstrate that our approach provides better average stream bit-rate than the earlier response time based method.

In this paper, we evaluate performances of 4K/2K video streaming using DASH-JS over 2.4/5GHz WiFi networks. We confirmed that fluctuation of reception quality was mainly caused by the minimum representation rate of MPEG-DASH contents and overlapped ON cycles for packet transmission.

Effective use of various wireless network interfaces is expected. In this paper, we evaluate Quality of Service characteristics of mobile 4K video using a radio communication quality map. We collect radio communication quality (e.g. throughput) by smartphones and visualize the quality in a heat map. Using this map, we select two routes that are expected to bring the highest and the lowest communication quality. We then evaluate QoS characteristic of 4K video streaming using MPEG-DASH on these two routes. The result shows that a good route user can achieve 1.5 times higher communication quality than a bad route user.

In this paper, we propose a method to control H.265/HEVC encoding for 8K UHDTV moving pictures by detecting amount or complexity of object motions. In 8K video, which has very high spatial resolution, motion has a big influence on encoding efficiency and processing time. The proposed method estimates motion features by external process which uses local feature points matching between two frames, selects an optimal prediction mode and determines search ranges of motion vectors. Experiments show we can detect motion complexity of 8K movies by using local feature matching between frames and we can select optimal configurations of encoding. By our method, we achieved highly efficient and low computation encoding.

本稿では,移動端末を対象とした快適経路ナビゲーションのコンテンツ先回り配信およびオフローディングシステムへの応用を検討した.コンテンツ先回り配信システムで提案されている配信スケジューリングを経路探索手法へ適応することで,安定した品質でビデオ配信を実現する快適経路の提案を行う.非常に単純なグラフ上で性能評価を行ったところ,ホットスポットのみで通信することから,セルラー配信時と比較して,大幅に通信時間を抑えることができ,端末の省電力効果が期待できることが分かった.

本稿では,スマートフォン上でMPEG-DASHを活用し,様々な無線規格を用いて4K映像配信時におけるQuality of Serviceの評価を行うことを目的としている.Long Term Evolution(LTE)や公衆無線LANサービスを利用し,周辺ユーザ数の影響,配信サーバへの同時接続端末数の影響,ユーザの状態(e.g.静止,徒歩,地下鉄移動)の影響を比較評価した.その結果,最も通信品質が良かったものは,5GHz帯のWi-Fiを利用時で,次にLTEであった.2.4GHz帯のWi-Fiは,最も品質が悪く,これは,周辺のAPの乱立による干渉が主な原因である.本実験で活用した4K映像では,2.4GHz帯のWi-Fi利用時を除き,選択されたビデオの品質も高く,再生フリーズも無かった.地下鉄乗車中においても,LTEを利用することで安定した再生が望めた.

今後、モバイル環境において大容量動画像コンテンツを中心にトラヒックが爆発的に増大する,膨大なトラヒックを処理するために、ネットワークとしてCDN (Contents Delivery Network)を経てネットワークのクラウド化を目的としてSDN (Software Defined Network)により仮想化ネットワークの実現を図ってきた.一方で、従来のIPネットワークのIPアドレスに依存しないネットワークとしてICN (Information Centric Network)が発表され、IPネットワークに対する優位性の検討が進められている.本論文ではICNを用いた列車内のコンテンツ配信システムとしてICNを用いて実証試験を行った結果について報告する.本研究は総務省SCOPE「先進的アプリケーション」の委託事業の協力により実証された成果である.

交通機関を利用しているユーザを対象とした高信頼ビデオストリーミングのためのバッファリング戦略手法の提案をする.本手法は,長期的スループット予測技術およびバッファ補充保証技術で構成される.長期的スループット予測技術を利用して,通信品質の変動を予測し,それに応じて確保する受信バッファ量を適応変化させるものである.その際,理論モデルにより,バッファ補充タイミングを算出することで,バッファ補充の保証を行う.性能評価として,実機評価したところ,通信品質が劣化したとしても,途切れない再生の実現を確認した.

We evaluated various characteristics of 4K video streaming using DASH-JS over LTE, public Wi-Fi services and WiMAX. We did experiments while stopping, walking, and riding the train. We found that a mobile user could enjoy smooth video streaming by using these networks even 4K video.

筆者らは,近年のモバイルトラフィックの爆発的な増加に対処するために,交通機関を活用したコンテンツ先回り配信システムを提案している.本稿では,その性能評価をするために,フィールド実験として,試験線を用いた実車両実験を行った.従来のCDNを想定したセルラーによる配信と比較し,先回り配信システムの有効性を明らかにした.

筆者らは，近年のモバイルトラフイックの爆発的な増加に対処するために，交通機関を活用したコンテンツ先回り配信システムを提案している．本稿では，その性能評価をするために，フィールド実験として，試験線を用いた実車両実験を行った．従来の CDN を想定したセルラーによる配信と比較し，先回り配信システムの有効性を明らかにした．We propose proactive content caching scheme utilizing transportation systems for robust video streaming with efficient wireless resources usage. In this paper, we develop a prototype system and evaluate its performances by using actual train vehicles and computer simulations. Evaluations conclude that our system can achieve higher-speed and highly-reliable video streaming rather than traditional CDN scenarios using cellular networks.

筆者らは，近年のモバイルトラフイックの爆発的な増加に対処するために，交通機関を活用したコンテンツ先回り配信システムを提案している．本稿では，その性能評価をするために，フィールド実験として，試験線を用いた実車両実験を行った．従来の CDN を想定したセルラーによる配信と比較し，先回り配信システムの有効性を明らかにした．We propose proactive content caching scheme utilizing transportation systems for robust video streaming with efficient wireless resources usage. In this paper, we develop a prototype system and evaluate its performances by using actual train vehicles and computer simulations. Evaluations conclude that our system can achieve higher-speed and highly-reliable video streaming rather than traditional CDN scenarios using cellular networks.

近年、移動しながら映像コンテンツを閲覧する機会が増えており、2017年には映像コンテンツがモバイルトラヒックの約7割を占めるとも予想されている。筆者らはこのモバイルビデオの省電力化に向けた検討を進めているが、既存のSDKを用いた評価では物理層までカスタマイズした評価は難しい。そこで本稿では、ソフトウェア無線の研究開発ツールとして広く使われているGNU Radioを活用し、変復調方式などを切り替えた場合のVLCストリーミングの特性評価を試みた。

筆者らは,交通機関を活用したコンテンツ先回り配信アプリケーションの提案を行っている.本稿では,その一つの重要要素である配信スケジューリング,スマートスケジューラの提案を行う。スマートスケジューラでは,中継点に先回り配信するコンテンツの品質,量,配信場所,配信タイミングを決めるものである.スマートスケジューラの性能評価および実機検証を行った結果,中継点におけるコンテンツサーバ間,列車サーバ間の帯域が十分に確保されている場合,スマートスケジューラは有効に機能していることが分かった.

我々は,高品質な通信を実現する快適経路の解析モデル提案をこれまで行ってきた.本稿では,より現実的な環境下として,HetNet環境下への拡張を行う.提案するモデルの検証として,実機実験との比較を行うため,実験フィールドである高田馬場における公衆無線LANスポットの評価を行った.具体的な実機検証では,移動経路,寄り道スポット数を変化させた実験シナリオを複数用意し,モデル検証を行った.その結果,いずれの実験シナリオにおいてもモデル精度は非常に良いことが分かった.

セルラトラヒック量は,ユーザ行動の影響により,日中のオフィス街と住宅地のように,空間的差異が生じる.本論文では,トラヒックの空間的不均一性を利用することで,セルラアップリンクトラヒックをオフロードするRoCNet(Robust Cellular Network)を提案する.RoCNetでは,蓄積運搬転送経路制御方式を用いて,トラヒック集中エリアにおいて生成されたトラヒックを低トラヒックエリアへオフロードする.また,パーティクルフィルタを用いることで,移動端末ごとに,過去の状態履歴に基づいて低トラヒックエリアへの移動可能性を予測し,移動可能性の高い移動端末へBluetooth等の近距離無線を用いて遅延を許容する生成データを転送することで,オフロード効果を高める.シミュレーションの結果,RoCNetを用いることで,トラヒック集中エリアから周囲の低トラヒックエリアへのトラヒック分散・平滑化が行われ,高トラヒックエリアの基地局の最繁時トラヒック量を約2割抑制できることを確認した.

画像の高能率符号化方式では、量子化に伴う情報量の削減とともに画質劣化が起こる。我々は過去に圧縮動画像の復号画像のみからPSNRの推定を行うNR (Non-Reference)型のブラインド推定を提案してきた。本稿ではその応用として、原動画像を入力として、H./264/AVCで圧縮した際のPSNRを予測する手法の提案を行う。その結果、80%以上の動画像フレームが予測誤差5dB以内でPSNRが予測可能であることが分かった。

近年,立体映像と多視点映像の標準化に注目が集まっている.通常,それらには動画像と奥行き情報をセットで用いる.また,Kinectや3Dカメラに代表される奥行き検出デバイスの普及が進んでおり,奥行き情報の取得が飛躍的に簡素化されつつある.今回我々は奥行き情報を用いてH.265/HEVCの2次元的な符号化画質改善を行う手法を提案した.具体的には,奥行き情報を用いて動画像の前景・背景分離を行い,その情報をもとに前景領域の位置検出をし,画質改善を行った.また,奥行き情報から動画像の境界領域の検出も行い,H.265/HEVCの精度改善を実現した.

一般に画像の客観画質評価の指標としてPSNR (Peak Signal-to-Noise Ratio)が用いられ、その計測には原画を用意する必要がある。しかし、YouTubeやニコニコ動画などのCGM (Consumer Generated Media)映像では、原画を用意することが困難な場合が多い。筆者らは原動画を用いず、かつビットストリームも解析せず、復号後の圧縮動画のみを用いてPSNRを推定するブラインド型PSNR推定の検討を行っている。これまでの検討では、推定精度を優先し動画の全フレームにおいて画質推定を行ってきたが、本稿では準リアルタイム処理を目指した高速化の検討を行った。全フレームにおいてブラインドPSNR推定を行う従来手法に対して、本稿では時間方向の情報を活用することで一定間隔のブラインドPSNR推定結果から全フレームの画質推定を行う高速化手法を検討する。

本論文では、全方位車載カメラ画像におけるイメージモザイキングの高速化について述べる。本論文における全方位車載カメラ画像とは、車の前方、後方、左右2つずつの計6台のカメラによって撮影される、走行中の周辺環境の画像群である。本画像群により走行中の全周囲画像を作成するため、イメージモザイキングを行いたい。筆者らは、モザイキングの処理フローを、画像ペアからのカメラ行列推定と、カメラ行列を用いた画像のモザイキング合成の2つの処理に分離し、相互に処理を独立させた。また、前地点までに推定したカメラ行列を後の画像に用いることで、推定処理を省略し、処理時間の短縮を試みた。さらに、推定が成功した地点でのカメラ行列を用いることで、カメラ行列が推定できなかった画像ペアに対して画像合成の実験を行い、結果を確認した。

Improving Quality of Service (QoS) in wireless networks is important and necessary for mobile users. We have previously proposed Comfort Route (CR) Navigation, which navigates users to their destinations using high QoS communication areas, such as Wi-Fi APs, rather than the geographical Shortest Route (SR). In this paper, we employ an analytical model to estimate the CR gain in a theoretical manner which assumes that available cellular and Wi-Fi throughputs are uniform within their coverage. The CR gain is computed by using basic parameters, including wireless network bandwidth and transmission time. To validate our model, we compare simulation results and real observation. These results conclude that the CR gain could estimate by using our analytical model.

To improve Quality of Service in wireless networks while mobile users travel, we introduce Comfort Route Navigation (CRN) which is a navigation system based on a user centric mobility management for next generation wireless networks. CRN provides an optimal route that satisfies user needs, such as obtaining maximum wireless resources. To achieve CRN, we construct an access point (AP) map of Shinjuku city. To reflect quality of APs in this map, we evaluate throughputs at seven public Wi-Fi spots. Based on these observations, we evaluate performance of our CRN via computer simulations and in a real environment. These evaluations conduct that a CRN user could obtain higher communication quality rather than a Shortest Route (SR) user. These evaluations also conclude that CRN gain depends heavily on the quality and location of the best broadband spot.

In recent years, video sequence with a depth map can be taken by a camera with RGB-D (depth) sensor like Kinect. On the other hand, quality degradation at occluded regions is inevitable when conventional 2-D video coding is applied. In this paper, we propose usage of an occluded region detection supported by the depth map for improvement of 2-D image coding quality. Experiments show that further improvement of picture quality at the whole of frame and occluded regions is addressed for H.265/HEVC compressed streams. © 2014 IEEE.

It is important for multimedia mobile users to obtain sufficient resources and a stable capacity of resources. Network resources such as available bandwidth for mobile users, however, fluctuate depending on the level of background traffic. The deployment of several access points enables mobile users to access more network resources on the way to their destination by taking the longcut route rather than the shortcut route. These users pay an extra cost to take the longcut route but obtain greater gains in throughput over the shortcut route. Previous research has shown that the gain-to-cost ratio can be more than double for the longcut route. No considerations, however, have been made for the case in which resource fluctuation is due to network congestion. This paper investigates the quality of service (QoS) for a longcut route with fewer resources (real resources) compared with that for one with the total resources (ideal resources) in the longcut route calculation. The results show that on average, users have gains of approximately double in real resources than their gains in ideal resources. This finding indicates that users can greatly improve their gains if they can accurately predict the resource fluctuation.

The peak signal-to-noise ratio (PSNR) used as an index of image quality usually requires original images, but this is difficult for consumer generated content such as videos on YouTube. Therefore, we developed two blind PSNR estimation methods without bit-stream analysis in which multiple support vector machines are prepared to learn differently encoded images in PSNR
using an entire frame and dividing the frame into two areas. We confirmed that higher estimation accuracy is possible for the latter method against that using the entire frame.

With the increasing demand for video streaming in mobile environment, we have been tackling a research on throughput prediction based on the measurement history data in order to improve service quality[1][2]. Since throughput fluctuation occurs sharply in mobile environment, useless packet loss or congestion occurs when available network bandwidth is insufficient compared to the bit rate of video contents. Hence, it is possible to improve the quality of service by deducing the available throughput and selecting appropriate bit rate for the video contents. In this paper, we propose a throughput prediction method that considers mobile locations of a train and their throughput history and show promising experimental results.

Extending battery life for smartphones while using wireless networks is important. In this paper, we propose Comfort Route for Energy Saving (CRFES), which helps users extend the battery life of their smartphones by navigating to their destinations via Quality of Service (QoS)-compliant energy-efficient spots such as Wi-Fi spots. To create CRFES, we construct maps of Wi-Fi spots using logging software we developed for Android phones. In addition to locations, we record actual throughput and energy consumption observed in cellular/Wi-Fi networks. So comfort route navigation is done by choosing a route which minimizes energy consumption instead of choosing the shortest path. Evaluations are carried out by computer simulation and one real city map. The results conclude that CRFES can save the battery life by approximately 50% by spending 1.6 times time cost to reach their destinations on average.

Traffic sign recognition contributes to the safety of drivers and people around the car. The system analyzes the road ahead images taken by on-vehicle camera, but blurred and distorted images make it difficult to recognize correct traffic sign. In this paper, we propose a method that deforms from ellipses to perfect circles and applies template matching with distorted sign templates.

Height data is very important not only as an indicator of growth in low birth weight infants, newborn, infants (infants below), but also for early detection of disease such as dwarfism. In this paper, we consider height estimation by image processing that targets lying infants for the purpose of reducing the burden of infants, doctors and parents. Furthermore, we compare results of using a single image and multiple images obtained by moving pictures, and show the latter contributes to improvement.

Feature analysis of audio compression is necessary to achieve high accuracy in musical content recognition and content-based music information retrieval (MIR). Bit rate differences are expected to adversely affect musical content analysis and content-based MIR results because the frequency response might be changed by the encoding. In this paper, we specifically examine its effect on the chroma vector, which is a commonly used feature vector for music signal processing. We analyze sound qualities extracted from encoded music files with different bit rates and compare them with the chroma features of original songs obtained using datasets for chord recognition. © 2014 Springer International Publishing.

Providing robust content delivery along with efficient wireless resource usage is important for next generation wireless networks. To achieve this, we propose a proactive content caching scheme utilizing transportation systems, especially trains. In our system, we place content servers with CCN capability to every train and station. Segments of video contents are pre-cached by the station servers before trains arrive at stations. Trains receive the contents via high-speed wireless transport while they stop at the stations. We develop a prototype system based on IP and CCN Hybrid protocols. We evaluate its performance by field experiment and compare with traditional CDN scenarios using cellular networks. Evaluations conclude that our system can achieve high-speed and high-reliable video delivery without freezing.

近年,モバイル端末におけるバッテリー寿命を延ばすことが重要な課題となっている.すでに,我々は,携帯端末における省電力を実現する快適経路ナビゲーションの提案を行っている.本稿では,より詳細な性能を解析,さらに推定するために,消費電力量をメトリックに加えた快適経路の解析モデルを提案する.解析モデルによると,省電力効果は,セルラーおよびWi-Fiのスループット,セルラー,Wi-Fi通信および非通信時の消費電力量,快適経路および最短経路のセルラー利用率,快適経路の寄り道度合で決定されることが分かった.また,モデル検証をするために,シミュレーション評価結果と比較したところ,解析モデルの精度は非常に良いことが分かった.

筆者らは,ICN (Information Centric Network)によるエネルギー利用効率の高い情報ネットワークの実現を目指すGreenICN (Architecture and Applications of Green Information Centric Networking)の取り組みにおいて,効率的なビデオ配信及びキャッシュ技術の研究開発を実施している.ビデオ配信のエネルギー利用効率を高める一つの方法として,無駄なパケットロスを極力減らすことが考えられる.パケットロスは,ビデオ配信レートに比して実際の帯域が不足している場合に発生し得るため,実際の帯域を推測してビデオ配信のレート決定に活用することで,効率的なビデオ配信が可能となる.本発表では,その初期検討として,無線指標から推測されるスループットと,時間帯や接続先基地局に基づいて推測されるスループットの内,小さい方をスループット予測値するモバイルネットワーク性能の予測手法及び,手法検討のための事前実験について報告する.

近年,無線ネットワークにおけるQuality of Service (QoS)の向上が重要課題として挙げられ,すでに,我々は,移動端末において高品質な通信を実現する快適経路ナビゲーションの提案を行っている.本稿では,より詳細に性能解析をするため,また,快適経路の性能を推測するために,快適経路の解析モデルを提案する.解析モデルのモデル検証として,シミュレーション評価との比較を行ったところ,我々の解析モデルの精度は非常に良いことが分かった.解析モデルによると,快適経路の効果は,セルラー,Wi-Fiのスループット比率,最短経路のWi-Fiカバー率,快適経路のセルラー利用率,ユーザコストによって決定され,特に,セルラー,Wi-Fiのスループット比率の影響が大きいことが分かった.

近年,立体映像と多視点映像の標準化に注目が集まっている.通常,それらには動画像と奥行情報をセットで用いる.また,同時に動画像の符号化を行うときオクルージョン領域の画質劣化は避けられないものである.我々は奥行情報を利用して,オクルージョン領域の検出を行い,その情報からオクルージョン領域の画質改善を行う手法を提案する.

近年,マルウェアによる被害が増加しており,対策が必要となっている.本研究では,マルウェアによる感染の有無を判定することで感染拡大を防ぐことを目標としている.本稿では,正常時のトラヒックデータとマルウェアに感染した後のトラヒックデータの特徴量毎の時間的な変化の違いに着目し,その時間的な変化をN-gramとしてまとめて扱うことで識別を行う手法を提案する.トラヒックデータの識別方法としては,正常時のトラヒックデータのN-gramと感染した後のトラヒックデータのN-gramの確率密度の大小で識別を行った.確率密度の算出のために最近傍密度推定法を用いた.マルウェアに感染した後のトラヒックデータとしてはCCCDATAset2010,2011を用い,正常なトラヒックデータとしては,あるPCで発生させたアプリケーショントラヒックを用いた.

Video streaming is one of the most popular Internet applications for years. To improve content availability and scalability, CDN and P2P networking have been widely used. However, the current Internet architecture, which relies on a host-based communication, has no ability to leverage caching at the network level. Recently, Content Centric Networking (CCN)has been emerged as a new networking paradigm. The communication of CCN focuses on the named contents instead of the IP addresses. This paper provides a framework for enabling video streaming over CCN. We performed experiments using CCNx on the PlanetLab to explore how an overlay path and a chunk size affect the streaming quality. Furthermore, we study the deployment of MPEG-DASH to achieve adaptive bit-rate streaming over CCN, which reveals additional challenge.

ストリーミングは、かつてはRTP/UDPとRTSPを用いる方式が主流であり、HTTP(TCP)ストリーミングはファイアウォールを超えられない場合の代替手段として用いられていた。しかし、インターネットの広帯域化とCDN (Content DeliveryNetwork)の普及に伴い、HTTP/TCPを用いての配信が主流になりつつある。さらに、Appleが提案したHTTP Live Streamingや、ネットワークの環境に応じて、配信するストリーミングデータの質を動的に変化させる事ができるMPEG-DASHが注目を集めている。一方、スマートフォンやタブレットPCの普及に伴い、LTEネットワークが急速に普及・発展している。そのため、本研究では、LTEハンドオーバーが発生する状況下において、HTTP Live StreamingやMPEG-DASHを行った場合、どのような問題点や課題が発生するかシミュレーション実験を行い、調査した。

近年、Underwater Sensor Networks は海中の観測データを収集するネットワークとして注目を浴びている田。しかし、海中環境は音波通信における伝送遅延の増加、雑音などでの様々な影響を受けるために、データ伝送が厳しい環境となっている。 Underwater Sensor Networks においても陸上通信と同様にルーティングプロトコルが必要となる。しかし、海中での様々な影響を受け、データ伝送が厳しい環境であるUnderwater Sensor Networkでは既存のルーティングプロトコルが有効とは言い難い。本稿では、ルーティングプロトコルに着目し、ネットワーク性能向上を目指す。中断や切断が多発したり、大きな伝送遅延が生じたりする劣悪な通信環境、ネットワークのことをDTN(delay/disruption-tolerant networking)という。そこで、本稿ではDTN向けのルーティングプロトコルとして提案されているEpidemic Routing[2]をUnderwater Sensor Networks に使用し、AODV(Ad-hoc On-demand Distance Vector)との比較検討を行う。

近年、海中環境の調査を効率的に行うため、Underwater Sensor Network が注目を集めている。 UnderwaterSensor Network では、AUV(自律型無人潜水機)と呼ばれる移動式のロボットにセンサ端末とモデムを搭載レ|青報の収集を行うが、その際エネルギーの消費を抑えることが重要となる。本研究では既存の研究を引用しつつ、音波モデムによるデータ送信とAUVの移動のそれぞれについて、調査に必要な消費エネルギーの定式化を行う。また、ネットワークシミュレータNS-3を用いてシミュレーション実験を行い、消費エネルギーを削減する方法について検討していく。

ネットワークのブロードバンド化に伴う様々なインターネットサービスが展開されている昨今,サービスを利用するユーザの増加がもたらすネットワーク全体またはサービスを処理するサーバへの負荷が高まることによるサービス品質の低下が問題視されている.これらの問題を解決する手法としてQoS技術が注目されており,ネットワーク設計において活発に用いられている.このQoS技術の中では,ネットワークを待ち行列としてしばしばモデル化する.近年,重要なテーマの一つとしてシック(遅延変動)を予測・モデル化することが挙げられる.平均待ち時間を考慮に入れた様々な解析手法が提案されているが,より高い精度を求めるために分散を考慮する手法を本稿では提案する.非常に解析が困難になることから,最も基本的なモデルであるFIFO (First-In-First-out)規律に従ったシステムを対象とし,GI/GI/1システムにおける分散の近似式を提案する.導出された近似式はシミュレーションによって精度を検証する.また,実トラフィックに於ける待ち時間の平均と分散を計測し,提案近似式が実環境に対してどの程度有効近似を示すかを検討する.

近年,Wi-Fiスポットが非常に混んでいると言われており,混雑度状況を可視化することが望まれている.本稿では,混雑度可視化を行うために,スループットとアクティブユーザ数の関係性を実機実験に基づき単純な近似式で表現し,またアクティブユーザ数の推移をモデル化する.また,混雑度の指標として利用するために,スループットだけでなく,アクティブユーザ数の推移によって期待されるスループット変動率の算出を試みる.また,混雑度の定義として,スループットとスループットの変動率を利用して4段階に分類し,精度評価を行った.今回の条件下では,70〜90%という比較的高い精度で混雑度推定が可能となることがわかった.

PSNRの計測には原画を用意する必要があるが、CGM映像では原画を用意することが困難な場合が多い。そこで筆者らはPSNRを変化させた多数の圧縮画像をSVMに学習させ、さらに顕著性マップを用いて画像を顕著領域と非顕著領域に分割することで推定精度の改善を図るブラインド画質推定に関する検討を進めてきた。しかし、顕著性マップでは、全体的に平坦な画像もしくは複雑な画像であっても相対的に顕著性領域の検出を試みるため、適切な分割が行われず精度が悪くなることがあった。そこで本稿では、動画において各フレームをACレベルを用いて2つに分割し、領域ごとにブラインドPSNR推定を行うことを試みる。これによって、顕著性マップによる分割が失敗する場合でも、安定したPSNR推定を行えることを確認した。

画像の高能率符号化方式では、低ヒットレート時にブロック処理に起因したブロック歪が再生画像中にしばしば発生する。圧縮画像の客観画質評価尺度としてPSNR、SSIMがよく用いられているが、これらはブロック歪に影響された画質劣化を十分に評価できるものではない。一方、復号画像におけるフロック歪の影響を評価する様々な手法が提案されているが、現時点で確立された手法があるとは言い難い。そこで本稿ではMPEG-2圧縮画像を対象に収集した多数の主観評価結果に基づき、従来手法で求めた各種の画像特徴量とブロック歪みの主観評価結果の相関性を評価する。また、その結果の応用として、圧縮画像のみからブロック歪みの影響量を推定するNR型画質評価手法についても検討を行う。

画像超解像品質は,雑音成分により劣化する.しかし,雑音成分は画像表現や解像感の向上に重要であるため,除去すれば良いというものではない.そこで我々は,原画像を信号成分と雑音成分に分離して各々超解像した後,合成を行うことを特徴とする方法を提案する.信号成分の超解像では,信号成分と,その多重解像度分解成分間のレジストレーションを行い,その超解像パラメータを最適化する.

本稿では,携帯端末に搭載されている地磁気センサ,加速度センサ,GPSから取得できるセンサデータを用いることによる,徒歩,電車,バス,自動車といったユーザの移動状態を識別する手法について述べる.各移動状態において携帯端末から取得したセンサデータを調査した結果,移動状態毎に差異が見られることを確認した.特に,電車における磁気量や,バス・自動車の各運転状況における加速度に固有の特徴が顕著に出現することが判明した.ここで,平均や標準偏差などの統計量や,パワースペクトルなどの特徴量が移動状態の識別に有効であるか検討を行った.

本稿では,携帯端末に搭載されている地磁気センサ,加速度センサ,GPSから取得できるセンサデータを用いることによる,徒歩,電車,バス,自動車といったユーザの移動状態を識別する手法について述べる.各移動状態において携帯端末から取得したセンサデータを調査した結果,移動状態毎に差異が見られることを確認した.特に,電車における磁気量や,バス・自動車の各運転状況における加速度に固有の特徴が顕著に出現することが判明した.ここで,平均や標準偏差などの統計量や,パワースペクトルなどの特徴量が移動状態の識別に有効であるか検討を行った.

Three-dimensional (3D) video service is expected to be introduced as a next-generation television service. Stereoscopic video is composed of two 2D video signals for the left and right views, and these 2D video signals are encoded. Video quality between the left and right views is not always consistent because, for example, each view is encoded at a different bit rate. As a result, the video quality difference between the left and right views degrades the quality of stereoscopic video. However, these characteristics have not been thoroughly studied or modeled. Therefore, it is necessary to better understand how the video quality difference affects stereoscopic video quality and to model the video quality characteristics. To do that, we conducted subjective quality assessments to derive subjective video quality characteristics. The characteristics showed that 3D video quality was affected by the difference in video quality between the left and right views, and that when the difference was small, 3D video quality correlated with the highest 2D video quality of the two views. We modeled these characteristics as a subjective quality metric using a training data set. Finally, we verified the performance of our proposed model by applying it to unknown data sets.

In Peer-to-Peer (P2P) networks, an incentive mechanism is a necessary component to deal with the free-riding behavior. The challenge is that direct reciprocal incentives
e.g., tit-for-tat, which consider the cooperation of peers in a pair-wise manner, are not suited with P2P streaming. In this paper, we propose a new service differentiation mechanism to provide a redistribution incentive for P2P streaming in a hybrid overlay network. The contribution of a peer can be measured from the number of video sub-streams that it uploads to other peers. By sending one request message, the number of sub-streams that each peer can retrieve is varied by its contribution level. An altruistic peer thus has to send less request messages and will experience smoother video quality than a selfish peer. Through simulations, we demonstrate that our solution can provide service differentiation among peers with better streaming quality than the tit-for-tat scheme.

Network resources for mobile users such as available bandwidth fluctuate depending on the level of background traffic. The deployment of many access points enables mobile users to access more network resources on their way to their destination by taking the longcut route rather than the shortcut route. These users pay some extra cost to take the longcut but obtain greater gains, such as in throughput, over the shortcut route. Previous research has shown that the gain-to-cost ratio can be more than double for the longcut route. No considerations, however, have been made for the case in which resource fluctuation is due to network congestion. This paper investigates the Quality of Service (QoS) for a longcut route with fewer resources (real resources) compared with that for one with the total resources (ideal resources) in the longcut route calculation. The results show that, on average, users have gains of approximately twice in real resources of their gains in ideal resources. This finding indicates that users can greatly improve their gains if they could accurately predict the resource fluctuation. Copyright © 2013 ACM.

This paper presents an RTT-fair TCP congestion control using ACK interval measurement and extends the approach to have inter-protocol friendliness, especially with CUBIC TCP. In the previous RTT-fair TCP congestion control, including ours, estimation of RTTs of a competing flow had been a problem. We try to solve this problem by measuring ACK arrival intervals, which are observable parameters by an end host. This approach enables estimation of congestion window behaviors, in addition to RTTs, of a competing flow. We then extend our congestion control to have friendliness to CUBIC-TCP, in addition to classical TCP-Reno, in an adaptive manner. Extensive experiment results are shown for simulations and implementations and effectiveness of our approach is confirmed.

This paper presents a TCP differentiation based on TCP version identification and IEEE 802.11e EDCA for low delay multimedia streaming over wireless LAN. It has been known that delay-based TCP can achieve low delay transport as long as it does not compete with loss-based TCP. However, when competition happens, it seriously decreases its rate by itself due to RTT increase. In order to alleviate this problem, we consider combination of TCP version identification and prioritized transport of delay-based TCP by EDCA at an access point. We evaluate this approach by simulations and implementations, and confirm its effectiveness. © 2013 IEEE.

In vehicle-to-vehicle communication, a message named BSM (Basic Safety Message) has a major role to inform a driver about surrounding condition. A vehicle periodically sends the BSM which includes the information of itself. Traffic load issues caused by BSMs would often be raised on heavily congested roads. Therefore, BSM congestion controls to avoid traffic congestion are challenging. This paper proposes a new BSM congestion control named BRAEVE. BRAEVE adapts its BSM generation rate according to the number of neighbor vehicles in communication range and then controls overall network traffic load. Our simulation evaluation shows that BRAEVE provides more uniform recognition of surrounding vehicles than an existing method, which brings much better safety assurance to a driver.

Objective picture quality measures cannot estimate the effect of blocking artifacts caused by video compression sufficiently. In this paper, we apply cepstrum analysis to quantify the blocking artifacts. We show experimental results for some test sequences using different coding schemes and prove effectiveness of our approach.

Quality of image super-resolution (SR) degrades by noise component. However it should not be eliminated because it is important for high definition impression for digital cinema. Therefore we propose an image SR method by synthesis of resolution-enhanced signal and noise components respectively after dividing an original image into signal and noise components. The signal component is resolution-enhanced using image registration between the signal component and its wavelet multi-scale components with resolution-enhanced parameter optimization. © 2013 IEEE.

Most network applications nowadays are related to the distribution of contents from a source to a number of users. Several developments such as CDN and P2P networking have been used to provide access to contents at the network edges. However, they have no ability to leverage caching at the network level and still rely on a host-based communication model. Recently, Content Centric Networking (CCN) has been emerged as a new networking paradigm. In CCN, the communication focuses on the named content instead of the IP address. This paper provides a framework for enabling video streaming over CCN. We performed experiments using CCNx on the PlanetLab to explore the effects of an overlay delivery path and a chunk size to the streaming quality. Furthermore, we study the deployment of MPEG-DASH to achieve adaptive bit-rate streaming over CCN, which reveals additional challenge. © 2013 IEEE.

We present a robust cellular network (RoCNet) that combines a cellular and an opportunistic networks for spatial uplink mobile data offloading, which focuses on the spatial difference of the traffic load among areas (e. g., business district and residential area in the daytime). RoCNet realizes the spatial data offload by leveraging the store-carry-forward routing mechanism. In the area where traffic load is high, delay-tolerant data originated from a mobile terminal is directly forwarded to a nearby terminal using Bluetooth or wireless LAN instead of being transmitted to a congested cellular base station. When the data is carried by the nearby terminal to other area where the traffic load is low, the data is forwarded to a cellular base station. To enhance the offload effect, it is necessary for data to be forwarded to a terminal that moves to a low traffic load area. In this paper, we use the particle filter to predict user behavior. Before forwarding data between mobile terminals, the terminals exchange prediction results and decide whether or not the data should be forwarded. We conducted a computer simulation whose result shows RoCNet can spatially offload uplink traffic in a traffic concentration area to non-congested areas. As a result, RoCNet can suppress peak traffic by about 20 percent in a traffic-congested base station by distributing traffic to vicinity base stations.

近年,スマートフォンやタブレット端末の普及,多様化により,今やスマートフォンの存在は人々の生活において無くてはならないものとなりつつあり,スマートフォンのバッテリー残量に非常に敏感になっている.このような背景のもと,我々は,すでに発表している最適寄り道経路に,端末の消費電力量をメトリックに加えたComfort Route for Energy Saving (CRFES) を提案する.CRFESでは,獲得通信量と消費電力量を両立した経路であり,Wi-Fi通信を利用することで,目的地へ向かうまでの所要時間が長くなるが,大きな省電力につながることがわかった.

近年のインターネット通信の高速化や利用できるサービスの多様化に伴い,トラヒック量の変動が著しい.トラヒックの混雑を回避するため,ネットワークを流れるトラヒック量の予測は,ネットワークを運用する上で非常に有益な情報であり,ネットワーク管理者が,事前にネットワーク上のトラヒック量を把握・制御することで,通信品質の劣化を防ぐことが望ましい.このため,筆者らは従来研究において十分な検討が行われていない,気象情報やカレンダ情報を始めとした外的要因に着目したトラヒック予測手法について検討している.本稿では,Webサービスへのアクセスログデータを評価対象として,ロジスティック回帰分析法を用いたトラヒック予測手法を提案すると共に,従来手法の一つである重回帰分析法との比較により,本手法の有効性を示す.

筆者らは以前にSIFTの平滑化処理にBilateral Filterを導入することで,精度面から性能を改善する手法を提案した.本論文では,本手法の応用例として一般物体認識への適用を検討する.従来手法との比較として,対応点探索や識別率の向上が確認できた.また,特徴点を与えるしきい値と識別率,特徴点数との関係も示す.

ビット深度スケーラブル動画像符号化方式とは,10bit深度などのHDR動画を,HDR環境・従来環境へ,冗長性を廃し同時に符号化する手法である.本論文では,ヒストグラムを補間する非線形フィルタを用い,レイヤ間予測の性能改善を行う手法を提案する.

近年,無線携帯端末,無線インフラの普及により,街中,移動中にインターネットサービスを受ける需要が高まっている.本稿では,無線LAN通信を効率良く利用するための最適寄り道経路手法を,従来の理論モデルによる評価の替わりに,実環境の実測データに基づくスループット特性モデルを用いて評価を行った.実環境の実測データとして,公衆無線LANサービスにおける実スループット特性を評価した.さらに、その結果を利用して,新宿を想定した評価マップにおける最適寄り道経路をシミュレーション評価,実機評価,両面において,最適寄り道経路の有効性を示した.シミュレーション評価,実機評価ともに,約15倍もの大きな寄り道効果が得られることがわかった.

従来のTCPは有線向けに設計されているためにウィンドウサイズを積極的に増加させる。そのため、無線マルチホップアドホックネットワークではMAC層でのコンテンションやパケット衝突などを引き起こし、性能低下に繋がる。そこで、TCP Vegasをベースにしたウィンドウサイズの増加幅・増加回数を抑制する輻輳制御方式が提案されている。低遅延なこの方式は、ノード配置がシンプルな環境では効果的に作用する。しかし、現実的で複雑な環境では効果的に作用しないことがあるため、本研究ではパケットロス後の対応を改良することで、改善を目指す。また、無線マルチホップネットワークにおけるTCPスループットの理論上限を参照しつつ、シミュレーションによって提案方式の有効性を示す。

本稿では、TCPバージョン識別とEDCA制御を併用したTCP差別化手法について報告する。TCPの輻輳制御の中で、遅延ベース手法のVegasは、RTTを基に輻輳制御を行うことで、RTTを抑えることができる。そのため、単独では無駄のない通信が可能である。しかし、Vegasは他のTCPと競合した際に、帯域を奪われてしまうといった親和性の問題がある。この親和性を改善するための1つの方法として、既にTCP差別化を提案し、シミュレーションにおいては評価を行った。本稿では、実機実験と実機に環境を合わせたシミュレーションによって、実環境においてのTCP差別化の効果を明らかにする。

近年、海中の環境調査を目的として、海中にSensor Networkを構成するというUnderwater Sensor Networkの研究が盛んに行われている。Undewater Sensor Nctworkでは、AUV(自律型無人潜水機)と呼ばれる移動式のロボットにセンサー端末とモデムを搭載し情報の収集を行う。その際エネルギー消費を抑えることは非常に重要である。本研究では、データ送信時に消費されるエネルギーとAUVの移動で消費されるエネルギーについて、数式とシミュレーションの両方を用いて比較検討を行う。

近年、Underwater Acoustic Sensor Networks(UWASNs)は海中の観測データを収集するネットワークとして注目を浴びている。しかし、海中における通信は利用帯域の制限、音波通信における伝搬時間の増加、様々な雑音など、厳しい環境において行われる。そのため、既存の陸地無線とは異なるUWASNsに適したデータリンク層やネットワーク層プロトコルの研究が多くなされている。本稿では、MACプロトコルでのデータ通信の信頼性を向上させるための自動再送制御(ARQ)、およびアクセス制御に着目した。ARQとして、selectiveとContinuous ARQとを組み合わせた手法を提案し、アクセス制御にはIEEE 802.11のCSMA/CAをUWASNs仕様に再設計を行い、新たなNAV(Network Allocation Vector)とCW(Contention Window)の取り決めであるSynchronous NAVおよびLinear CWを提案する。それらをMACプロトコルとして、シミュレーション通じてネットワーク性能の評価を行う。

近年,ITS (Intelligent Transportation System)の技術が急速に発展してきている.ITSにおける車車間通信では,BSM(Basic Safety Message)と呼ばれるメッセージに自車の情報を載せて,定期的に周辺車両に知らせることにより,危険な状況を事前にドライバーに伝える重要な役割を果たす.高速道路等での渋滞した道路状況では車両が密に存在するため,BSMによるトラフィック負荷が問題となってくる.そこで,BSMの送信頻度(レート)や送信パワーを変化させることにより,輻輳を回避する手法が近年研究されている.本研究ではBSMのレート制御に着目し,従来手法に関して,より現実に近い状況下において問題があることを指摘すると共に,解決策とその検討をシミュレーション実験により行った.

TCPの輻輳制御方式において,他のコネクションとの公平性を高めるためには,ネットワークの状態を推定情報を得ることが有効である.ネットワーク状況を推定するためには,中継ノードにおいて計測を行い端末にフィードバックを行う手法がある.しかしインターネットなど現状のネットワークにおいてTCPの輻輳制御が自律分散制御であることを考慮すると端末が直接取得可能な情報を利用して推定を行うことが望ましいと考えられる.よって本稿では,端末が直接取得可能なACKの到着間隔に着目してネットワークの状況の推定を行い,その情報を用いてRTT公平性の高い輻輳制御を行うことや新たに標準となりつつある制御方式へ親和性のある制御ができることを示す.

筆者らは,セルラに代表されるインフラネットワークと,DTN(Delay Tolerant Network)に代表される定期/不定期に到来する通信機会を有効利用して通信を実施するopportunisticネットワークとを組み合わせ,災害時通信と通常時のトラフィックオフロードを実現するロバストセルラネットワーク(RoCNet)の検討を実施している.RoCNetは,災害時にはDTNのような蓄積・運搬・転送による通信手段を提供する一方,通常時は局所的なトラフィック輻輳を緩和するために蓄積・運搬・転送の仕組みを利用する.本稿では,RoCNetの概要を紹介するとともに,RoCNetによりトラフィック集中エリアのアップリンクトラフィックを輻輳状態でないエリアに空間的にオフロード可能というシミュレーションによる初期評価の結果を示す.

本稿では,3軸加速度センサを用いた歩行状態(平地歩行,階段上り,階段下り)識別における音声処理で使われているケプストラム分析を応用することの有効性を示す.ケプストラムが歩行時の加速度の平均や分散等の統計量と比較して,識別精度が高くさらに経年変化や路面状態変化といった取得環境の影響を受けにくい特徴量であることを実験的に確認した.そして,その理由についてケプストラム分析により分離したスペクトル包絡の特徴を調べることにより考察した.

This paper focuses on RTT-fairness of multiple TCP flows over the Internet, and proposes a new TCP congestion control named "HRF (Hybrid RTT-Fair)-TCP". Today, it is a serious problem that the flows having smaller RTT utilize more bandwidth than others when multiple flows having different RTT values compete in the same network. This means that a user with longer RTT may not be able to obtain sufficient bandwidth by the current methods. This RTT fairness issue has been discussed in many TCP papers. An example is CR (Constant Rate) algorithm, which achieves RTT-fairness by multiplying the square of RTT value in its window increment phase against TCP-Reno. However, the method halves its windows size same as TCP-Reno when a packet loss is detected. This makes worse its efficiency in certain network cases. On the other hand, recent proposed TCP versions essentially require throughput efficiency and TCP-friendliness with TCP-Reno. Therefore, we try to keep these advantages in our TCP design in addition to RTT-fairness. In this paper, we make intuitive analytical models in which we separate resource utilization processes into two cases: utilization of bottleneck link capacity and that of buffer space at the bottleneck link router. These models take into account three characteristic algorithms (Reno, Constant Rate, Constant Increase) in window increment phase where a sender receives an acknowledgement successfully. Their validity is proved by both simulations and implementations. From these analyses, we propose HRF-TCP which switches two modes according to observed RTT values and achieves RTT fairness. Experiments are carried out to validate the proposed method. Finally, HRF-TCP outperforms conventional methods in RTT-fairness, efficiency and friendliness with TCP-Reno.

Previous our research has introduced an optimal longcut route problem and a number of solutions to this problem. However, to date, the literature has only presented evaluations based on a single theoretical radio propagation model. Because throughput characteristics derived from radio propagation models significantly affects longcut route performance, this study introduces three simple models by referring to actual measurement values. The results of our evaluation show that there are large differences in performance between the three models. The Non-Line-of-Sight model can have an improvement ratio approximately five times greater than that of the Line-of-Sight Indoor model and six times greater than that of the previous theoretical model. © 2012 IEEE.

In Peer-to-Peer (P2P) networks, a great number of peers are reluctant to contribute to the system. An incentive mechanism thus becomes a necessary component to motivate the cooperation of peers. Moreover, due to the continuous nature of data transmission, the P2P media streaming requires different incentive strategies from the P2P file-sharing. In this paper, we propose a new service differentiation mechanism to provide a redistribution incentive for P2P streaming in a hybrid overlay network. The contribution amount of a peer can be measured from the number of video sub-streams it transmits, instead of the quantity of transferred data. By sending one request message, the number of sub-streams that each peer can retrieve is varied by its contribution amount. An altruistic peer thus has to send less request messages and experiences smoother video quality than a selfish peer. Through simulations, we demonstrate that the proposed mechanism can provide service differentiation among peers according to their contribution amount. © 2012 IEEE.

The growing popularity of mobile data communication has been leading a growing lack of radio resource on a cellular network. Therefore, to avoid the data traffic from exceeding the capacity limit of a cellular base station (BS), cellular operators have been trying to install cellular femto cells (local BSs) or wireless LAN access points (APs) in restaurants, cafes, user's houses, etc., to divert data traffic from the cellular radio access link to the fixedline such as x-DSL and FTTH. However, the traffic levels extremely change in a day, even locally, due to the mobility of users. While peak data traffic at certain times of a day appears locally, cellular BSs as well as the femto cells and the APs would sit underutilized for the rest of the day. Thus, we have started designing a robust cellular network (RoCNet) that combines infrastructure-based network (e.g., cellular network) and opportunistic networking for spatial mobile data offloading, which focuses on the difference of the data traffic level among areas (e.g., business district and residential area). The RoCNet facilitates the data traffic offloading leveraging the storecarry-forward feature considering a degree of traffic congestion. In this paper, we give an overview of the RoCNet and show a simulation result. The result shows RoCNet can spatially offload uplink traffic in a traffic concentration area to non-congested areas. In other words, RoCNet can suppress peak traffic in a traffic-congested base station by distributing traffic to vicinity base stations. Copyright © 2012 ACM.

Bit-depth scalable coding method is an approach that generates multiplexed bit-streams that can be decoded to two video sequences, for Standard Dynamic Range (SDR) environment and for High Dynamic Range (HDR) environment. This paper presents a bit-depth scalable coding method that is considered as gradation restoration in inter-layer prediction process. Our proposed inter-layer prediction uses a histogram interpolation method that enables to generate more mid-gray brightness levels than traditional inverse tone mapping methods with one-to-one correspondence.

This paper presents a chord recognition method from music signals using chroma vectors and musical knowledge known as Doubly Nested Circle of Fifths (DNCOF). DNCOF represents the relationships of major and minor chords where the neighboring two triads are similar. We obtain a novel feature from chroma vectors by mapping them onto two-dimensional DNCOF coordinate, which we call DNCOF vectors. We expect that the DNCOF vectors can contribute to correcting false recognition obtained by the chroma vectors when their mapped positions are apart from one another in the DNCOF coordinate. In this research, we evaluated our proposal using the Beatles' datasets and showed its effectiveness.

Feature extraction of images can be applied to image matching, image searching, object recognition, image tracking etc. One of the effective methods to extract features of images is Scale-Invariant Feature Transform (SIFT) [1]. In this paper, we indicate problems of SIFT and propose a method to improve its performance by applying Bilateral Filter [2]. In addition, we implement is acceleration by GPGPU (general purpose GPU), apply this method to generic object recognition and perform a comparison experiment. We compare the proposed method with the original method using SIFT and confirm improvement of the identification rate by the proposed method.

We propose a novel wavelet domain image super-resolution method from digital cinema to ultrahigh definition television considering cinema noise component. The proposed method features that spatial resolution of an original image is expanded by synthesis of super-resolved signal and noise components respectively after dividing an original image into signal and noise components. Dividing noise component uses spatio-temporal wavelet decomposition based on frequency spectrum analysis of cinema noise. And super-resolution parameters are optimized by comparing size-reduced super-resolution images with an original image. Experimental results showed that a super-resolution image using the proposed method has a subjectively better appearance and an objectively better peak signal-to-noise ratio measurement than conventional methods.

This paper presents a music part detection method incorporating chroma vector analysis for use with music TV programs. Results show that envelopes of chroma components of music signals tend to have horizontal (i.e. temporal) correlation in time-frequency representation because music signals have a periodic chord sequences. Based on this fact, we analyze time series of chroma components and attempt to segment music parts in music TV programs from other parts. Experimental results show an F-measure of 0.78, which is better than that obtained using the previous method.

楽器演奏者にとって,練習環境の確保は重要である.しかし,ドラムについては楽器の特性上,奏者の好きな場所,時間に練習することは難しい.そういった問題を解決したインタフェースとして、Wii Musicなどが挙げられるが、打ち分け操作が直感的とはいえず、ドラム練習としては不十分であった。また、文献[2]において、ジャイロセンサを利用することで、打ち分けを直感的にする試みを行ったが、打ち分け精度に問題があった。そこで本稿では、加速度センサ・ジャイロセンサ・深度センサの三つのセンサを組み合わせる新しいドラムインタフェースを提案する。

The presence of free-riding peers has been known as a serious threat in the peer-to-peer networks as it may lead to the degradation of overall service quality. In this paper, we study a method to construct the push-based multicast overlay in a contribution-aware manner for P2P streaming. The idea is to construct the tree-shaped core overlay and define a set of auxiliary links to improve its resiliency. The auxiliary connections, which are constructed based on the service capacity and contribution, also alleviate the peers to deliver data using a multi-parent streaming approach. Simulations through ns2 demonstrate the efficiency of our solution.

In this paper, we introduce a hybrid approach for overlay construction in an application-layer multicast and its application to a distributed data center management. We combine the good features of a tree-based structure and a mesh-based data delivery to form ToMo, a two-layer hybrid overlay. Different from other approaches, the overlay is constructed in the fashion of a mesh layer over a tree layer. This structure helps to reduce the number of replicated packets at a source by allocating the source to multicast each packet to a specific group of child peers only. We also consider a simple application of our overlay structure to the failover for data centers. That is, the servers of distributed data centers are connected by a mesh, and the clients are hanging on the servers in a tree structure. Upon failure, the server can be switched over to a redundant server follows the defined mesh routes automatically. Simulations through ns2 demonstrate the efficiency of this solution.

To improve driving comfort and provide entertainment services, vehicular communication networks (VCDNs) have appeared as an emerging solution, which consists of road-side units (RSUs) and on-board units (OBUs) to distribute multimedia contents. However, as most of OBUs always request the stored contents in the RSUs, how to update the contents in these RSUs when the original changes at its original servers has become an important issue to be dealt with. This paper proposes a novel method to resolve the above problem. Firstly, based on the characteristics of peers and geographical information, we decide which replica of which content in RSUs should be updated when its original changes. Secondly, by comparing the delivery cost of wired and wireless transmission, we decide whether the updated content should be delivered from a fixed peer or other mobile peers. Lastly, the detailed algorithm is presented and summarized.

In this paper, we introduce a hybrid approach for overlay construction and data delivery in an application-layer multicast. We combine the strong points of a tree-based structure and a mesh-based data delivery to form ToMo, a two-layer hybrid overlay. We try to reduce the number of replicated packets at a source, and reduce an effect when slow connection peers are located near the source. The overlay is constructed in the fashion of a mesh layer over a tree layer. This structure allocates the source to multicast each piece of the packet to a specific group of child peers only. Different from other approaches, we employ only push-based data delivery in order to minimize the latency. The redundancy is avoided by defining a set of well-organized mesh connections. Furthermore, in our approach, the isolated peers affected by parent departure are not facing data loss during the rejoin process since they still receive data from their neighbors via mesh connections. Simulations through ns2 demonstrate the efficiency of this solution.

The hybrid Application Layer Multicast (ALM) has been shown its efficiency by leveraging the conventionally main structures of application layer multicast, tree-based and mesh-based. However, how to select the proper node to construct the overlay and how to establish the connection between any two nodes are still unsolved. Therefore, this paper is to design a novel construction algorithm for the hybrid AIM to resolve the above two issues. Firstly, by carrying out the analysis of nodes&apos; characteristics, all nodes are divided into groups and a node priority is proposed to select the super node within each group. Secondly, by using the selected super node, all nodes are hierarchically controlled and different kinds of connections are carried out in the AIM, where the connection between super node and other nodes is set to be a tree to enhance the efficient utilization of network resource while the connection between other normal nodes is decided to be a mesh to reduce the overhead. Simulation results show that the proposal outperforms other conventional methods.

With the development of broadband technologies, more and more multimedia contents such as video or bulk file are being distributed over the Internet. However, an important issue to be resolved is how to construct an efficient overlay to provide users with the acceptable throughput and delay, while receiving multimedia contents. Therefore, this paper designs a tree construction algorithm by distributing the layered steaming over the ALM in order to improve both throughput and user delay. Firstly, to improve the throughput, based on theory analysis, we define a layered degree, out/in-degree and the corresponding constraints to manage the layered streaming and nodes. Secondly, a novel method, called Bi-cast, is proposed to reduce user delay during data-transmission. Thirdly, by using the defined degrees and the Bi-cast, we present a tree construction algorithm. Both the simulation and the implementation are carried out, and the results show that our proposal can obtain better performance than other conventional methods.

This paper presents a hybrid congestion control algorithm supporting throughput efficiency, TCP friendliness and RTT fairness, in order to provide highly smart multimedia streaming over the Internet. Recent studies on TCP and TFRC are evolving to achieve better throughput efficiency and RTT fairness. Examples are Hybrid TCP congestion control, VTP and TCP-Libra, where the former two improve throughput efficiency of TCP and TFRC, respectively, and the last one improves RTT fairness of TCP. This paper integrates these ideas and propose a new congestion control algorithm, called Hybrid TFRC, for multimedia streaming. Experiments are carried out and effectiveness of our proposal is verified. ©2010 IEEE.

Service capacity and contribution of the peers have been proven to be the important factors for selecting a serving peer in the P2P overlay networks. In this paper, we study how to design a multi-parent overlay network in a contribution aware manner. Our idea is to utilize the outbound bandwidth of all overlay nodes efficiently. Different from other multisource streaming approaches, we employ push-based data delivery in order to minimize the latency. The core of our overlay is conceived in a tree structure rooted at a source. The auxiliary connections with redundancy avoidance capability are proposed to increase the resilience of the tree. Furthermore, we introduce an overlay joining algorithm where the parent whose degrees are exhausted responses the new peer with a list of potential peers having residual degrees so that the new peer can join the overlay with minimal requests. Simulations through ns2 demonstrate the efficiency of our solution. ©2010 IEEE.

With the development of recent vehicular communication technologies, distributing multimedia contents in the vehicular communication networks (VCNs) has become more and more popular, to provide conveniences and entertainment services during the time of driving. However, as multimedia contents are changed and updated dynamically, how to keep the consistency between the original and these replicas in VCNs is very important. Therefore, this paper designs a novel algorithm to control the consistency for the VCNs. In our proposal, after the analyses of the status of road-side units, on-board units and local geographical information, we divide all replicas into different groups, where one is necessary for update and the other are not. Then, we compare the cost to update replicas by using wireless and wired connection, and propose a method to make selection between them. The performance of our proposal is tested by simulation experiments. And the results show that our method can obtain the low delay successfully.

This paper proposes a new TCP congestion control supporting loss-fairness, which also contributes to improving RTT fairness. It has been well known that smaller RTT flows grab more bandwidth than larger ones when they compete on the shared link. This is called RTT unfairness and is caused by congestion control mechanism of the current TCP. To solve this problem, many protocols have been proposed. TCP-Libra is one of them, which formulates its window increase rate as a function of RTT and brings the same throughput irrespective of different RTT values. However, we indicate TCP-Libra's formulation holds only when the congestion window is halved upon packet losses similar to TCP-Reno. When we apply different (smarter) window control schemes, the formulation has to be modified accordingly. Therefore, in this paper, we consider a hybrid congestion control supporting RTT fairness which switches two modes according to buffering state at the bottleneck link. In decrement phase upon packet loss, it decreases window size to clear the buffering packets instead of halving the window size. Furthermore, we introduce the concept of loss-fairness for our window increase formulation, which is equivalent to RTTfairness. Experiments are carried out to validate the proposed method by both implementation experiments and simulations, and much better performances are achieved against conventional methods. ©2010 IEEE.

This paper presents a novel key estimation method of sound sources based on the music theory known as "circle of fifths". We firstly overview music theory and formulate the musical key analysis by vector operations. In detail, we separate music sources into small pieces and calculate FFT-based chroma vectors. They are converted to tonality vectors and COF (circle-of-fifth) vectors are calculated from the tonality vectors, which are mapped onto the circle of fifths coordinate. As a result, each music source can be represented by traces of COF vectors, which usually stay inside a single key region on the circle of fifths. Finally, HMM is applied to the traces of COF vectors in order to detect keys and their boundaries. Experiments using music databases are also carried out.

In order to achieve highly accurate content-based music information retrieval (MIR), it is necessary to compensate the various bit rates of encoded songs which are stored in the music collection, since the bit rate differences are expected to apply a negative effect to content-based MIR results. In this paper, we examine how the bit rate differences affect MIR results, propose methods to normalize MFCC features extracted from encoded files with various bit rates, and show their effects to stabilize MIR results.

With the growing use of dynamic content by mobile content distribution systems, how to manage dynamically changing files has become an important issue, since the cached replicas on different mobile sites must be updated if the originals have been changed. Therefore, this paper proposes a priority based selection method to enhance the efficient utilization of network resource and support the client mobility for Mobile Contents Delivery Network (M-CDN). On one hand, a consistency priority is calculated by analyzing the characteristics of mobile surrogates. If a given content which has been changed on its original node, only the replicas with the high consistency priority instead of all replicas are updated. On the other hand, an update priority is also proposed. If a replica has been selected for update, the latest version will be sent from the site decided by the update consistency. Simulation results show that the proposed new approach outperforms other conventional methods. © 2009 IEEE.

This paper presents TCP-Fusion supporting RTT (Round Trip Time) fairness in addition to throughput efficiency and friendliness to TCP-Reno. When multiple TCP flows having different RTT values compete, more bandwidth is unfairly allocated to the flow having smaller RTT. This means that a user with longer RTT may not be able to obtain sufficient bandwidth by the current methods. On the other hand, recent studies on the TCP congestion control to achieve RTT fairness and throughput efficiency are evolving actively. An example for RTT fairness is TCP-Libra and an example for throughput efficiency is Hybrid TCP congestion control. This paper focuses on Hybrid TCP (exploiting residual link capacity when TCP-Reno drops its rate) and improves its RTT fairness by incorporating the ideas of TCP-Libra (congestion window increase in proportion to square of RTT) to its loss mode. Experiments are carried out to validate the proposed method and much better performances in RTT fairness and throughput are provided against conventional methods.

In this paper, we introduce a new mesh-based approach for data delivery which is organized over multiple tree-shaped core routes. Given that both tree and mesh approaches for multicast overlay construction have their own strong points, we simply combine them. During session, the mesh connections are formed dynamically to facilitate the data delivery on tree-shaped core routes which may experience bandwidth bottlenecks. Unlike existing mesh-based approaches, our proposal employs push-based data delivery along mesh connection. A well-organized mesh construction scheme is defined to avoid data redundancies. We evaluated the proposal through ns-2 simulator. The simulation results demonstrate that our approach can provide higher average received quality and has acceptable control overhead when compared to a single tree method. We also show that, over a static overlay, the push-based data delivery on mesh can provide the received quality close to supposed pull-based data delivery method with less latency. Moreover, it has lower control overhead than the pull-based method when the peer number is large.

This paper is to design an integrated consistency-control algorithm for mobile contents delivery network (M-CDN) to enhance the efficient utilization of network resource and support the client mobility. Firstly, by carrying out an analysis of mobile surrogates&#039; characteristics, for a given content which has been changed on its original node, only a limited number of its replicas instead of all replicas are updated. Then, if a replica has been selected for update, the latest version will be sent from an algorithm-decided site instead of from its original server.

This paper presents a denoising method of intra-coded pictures using motion estimation and pixel shift. Firstly, we show that pixel-aligned mixture of distorted images which are spatially shifted and differently encoded brings reduction of quantization errors. We show that this effect can be formulated as a special case of Wiener-Hopf equation and independence of quantization errors affects the performance. We then consider its application to denoising of intra-coded pictures by using motion estimation and pixel shift. Experiments using actual image sequences verify that motion estimation is effective in moving regions, pixel shift is effective in static regions and favorable PSNR gains are achieved.

With the development whereby mobile content distribution systems can manage dynamically changing files, an important issue to be resolved is consistency management, which means the cached replicas on different mobile sites must be updated if the originals change. This paper is to design an integrated consistency-control algorithm for mobile contents delivery network (M-CDN) to enhance the efficient utilization of network resource and support the client mobility. Firstly, by carrying out an analysis of mobile surrogates' characteristics, for a given content which has been changed on its original node, only a limited number of its replicas instead of all replicas are updated. Secondly, if a replica has been selected for update, the latest version will be sent from an algorithm-decided site instead of from its original server. Simulation results show that the proposal outperforms other conventional methods.

This paper presents simple analytical models of hybrid TCP congestion controls, which switch loss-based mode and delay-based mode adaptively, and tries their performance tuning. We firstly present ideal behavior models of three kinds of TCP congestion controls (loss-based, delay-based and hybrid). We then give abstracted models of the actual hybrid TCP s and consider their performance tuning. Finally, experiments validate analytical expectations and effectiveness of the hybrid TCP.

This paper is to design an integrated consistency-control algorithm for mobile contents delivery network (M-CDN) to enhance the efficient utilization of network resource and support the client mobility. Firstly, by carrying out an analysis of mobile surrogates&#039; characteristics, for a given content which has been changed on its original node, only a limited number of its replicas instead of all replicas are updated. Then, if a replica has been selected for update, the latest version will be sent from an algorithm-decided site instead of from its original server.

This paper is to design an integrated consistency-control algorithm for mobile contents delivery network (M-CDN) to enhance the efficient utilization of network resource and support the client mobility. Firstly, by carrying out an analysis of mobile surrogates characteristics, for a given content which has been changed on its original node, only a limited number of its replicas instead of all replicas are updated. Then, if a replica has been selected for update, the latest version will be sent from an algorithm-decided site instead of from its original server.

Content delivery network improves end-user performance by replicating Web contents on a group of geographically distributed sites interconnected over the Internet. However, with the development whereby content distribution systems can manage dynamically changing files, an important issue to be resolved is consistency management, which means the cached replicas on different sites must be updated if the originals change. In this paper, based on the analytical formulation of object freshness, web access distribution and network topology, we derive a novel algorithm as follows: (1) For a given content which has been changed on its original server, only a limited number of its replicas instead of all replicas are updated. (2) After a replica has been selected for update, the latest version will be sent from an algorithm-decided site instead of from its original server. Simulation results verify that the proposed algorithm provides better consistency management than conventional methods with the reduced the old hit ratio and network traffic.

With the development whereby mobile content distribution systems can manage dynamically changing files, an important issue to be resolved is consistency management, which means the cached replicas on different mobile sites must be updated if the originals change. This paper is to design an integrated consistency-control algorithm for mobile contents delivery network (M-CDN) to enhance the efficient utilization of network resource and support the client mobility. Firstly, by carrying out an analysis of mobile surrogates&apos; characteristics, for a given content which has been changed on its original node, only a limited number of its replicas instead of all replicas are updated. Secondly, if a replica has been selected for update, the latest version will be sent from an algorithm-decided site instead of from its original server. Simulation results show that the proposal outperforms other conventional methods.

The purpose of this study is to construct a robust and stable overlay tree of ALM (Application Layer Multicast) for real-time video transmission. Firstly, we propose a proactive route maintenance which enables fast recovery of the overlay tree against node departures and failures. By forcing free node degrees for route backup, signaling overheads for route maintenance can be reduced in a scalable manner. Secondly, we improve performance of the proposed method by introducing layered video coding. Smooth layer management instead of coarse degree management contributes to reduction of the depth of overlay tree (i.e. delay) as well as to efficient bandwidth utilization. Thirdly, we introduce session records of each node into the overlay tree construction. Long-term history of user access records is expected to bring more robust overlay trees. Finally, we carried out extensive evaluation experiments. Simulations and implementations demonstrate that our methods lead to fast recovery of the overlay tree against node departures and reduction of the signaling overheads. Furthermore, introduction of layered video coding is proved to be efficient to reduce delays, to improve bandwidth utilization and to avoid severe degradation of picture quality. Final experiments show that incorporation of node stability reduces the number of nodes affected by parent node departures by promoting adequate nodes to upper layer of the ALM tree. ©2007 IEEE.

Content Delivery Overlays improves end-user performance by replicating Web contents on a group of geographically distributed sites interconnected over the Internet. However, with the development whereby overlay systems can manage dynamically changing flies, an important issue to be resolved is consistency management, which means the cached replicas on different sites must be updated if the originals change. In this paper, based on the analytical formulation of object freshness, web access distribution and network topology, we derive a novel algorithm as follows: (1) For a given content which has been changed on its original server, only a limited number of its replicas instead of all replicas are updated. (2) After a replica has been selected for update, the latest version will be sent from an algorithm-decided site instead of from its original server. Simulation results verify that the proposed algorithm provides much better consistency management than conventional methods with the reduced the old hit ratio and network traffic. © 2007 IEEE.

A robust replication algorithm is very important to retrieve Web objects from the Content delivery network (CDN), where both the network topology and Web access are changed dynamically. However, current replica strategies in CDN are to statically and repeatedly keep the complete replica of the original object on many content servers. This method is weak in adapting the dynamic network, resulting in too much consumed server space and a waste of the storage cost. It becomes more serious by replicating some large-sized objects such as streaming media, which are being distributed over the Internet,more and more. In this paper, we discuss a robust replication strategy for scalable video strew-ning in CDN to reduce user response and storage cost as follows: 1) For content servers, which layer of which video contents should be selected to replicate? 2) How to test the robustness under different network environments? Simulation results show that the proposed algorithm is robust and can provide much better performance than the conventional methods.

The purpose of this study is to maintain efficient backup routes for reconstructing overlay trees quickly. In most conventional methods, after a node leaves the trees, its child nodes start searching for the new parents. In this reactive approach, it takes a lot of time to find a new parent. In this paper, we propose a proactive approach to finding a next parent as the backup route node over the overlay tree before the current parent leaves. A proactive approach allows a node to find its new parent node immediately and switch to the backup route node smoothly. In our proposal, the structure of the overlay tree using a redundant degree can decide a backup route node without so much overhead. Simulations demonstrate our proactive approach can recover from node departures 2 times faster than reactive approaches, and can construct overlay trees with lower overheads than another proactive method. Additionally we carried out experiments over actual networks and their results support the effectiveness of our approach. We confirmed that our proposal achieved better streaming quality than conventional approaches. Copyright © 2006 The Institute of Electronics, Information and Communication Engineers.

We propose an on-demand multipath routing algorithm in a mobile ad hoc network for video transmission and evaluate its real world performance in video streaming application. There have been a number of multipath routing protocols extended from AODV which is a well-known single path routing protocol. Multipath routing protocols indicate good performance in the reduction of route discovery latency and unnecessary routing packets in simulations. We show that the route establishment using source route lists provided by us (Hu and Johnson, 2002) can reduce the route discovery latency, select stable routes automatically, and work well for live video streaming without limitation of the hop count based approaches. We evaluate this proposed method compared with the original AODV by using eight laptop PCs and demonstrate live streaming experiments.

The current video switching/composition scheme employed is fundamentally subject to a variable delay of 0-1 frame at each stage due to the use of a frame synchronizer. We can fix and minimize this variable delay by controlling the video sync phase of the distributed video sources at the start of the communication session if they are driven by either the same video sync frequency or independent but very accurate sync frequencies. Furthermore, delivering the necessary phase controlling signals between the terminal and the video switching or composing point can be relatively easy because of the widely available IP networks. Based on .these ideas, we determined the performance of an experimental system we constructed that used video phase synchronization of distributed video sources. We describe the system's architecture, discuss the sync source dependency of the system, and report the delay observed in existing and proposed schemes of video composition using real time hardware experiments. The results obtained indicate that the proposed video sync controlling system is effective and has practical applications for fixing and minimizing the video switching/composition delay.

The purpose of this study is to maintain efficient backup routes for reconstructing overlay trees quickly. In most conventional methods, after a node leaves the trees, its children start searching for a new parent. In this reactive approach, it takes a lot of time to find a new parent. In this paper, we propose a proactive approach to finding a new parent over the overlay trees before the current parent leaves. A proactive approach allows a node to find its new parent immediately and switch to the backup route smoothly. In our proposal, the structure of the overlay tree using a redundant degree can decide a new parent without so much overhead. Simulations demonstrate our proactive approach can recover from node departures 2 times faster than reactive approaches, and can construct overlay trees with lower overheads than another proactive method. Additionally we carried out experiments over actual networks and their results support the effectiveness of our approach. We confirmed that our proposal achieved better streaming quality than conventional approaches.

This paper presents a novel approach for efficient tree construction in ALM (Application Layer Multicast) using layered video coding. This system is suitable for the users in heterogeneous environment such as xDSL and CATV, where upload and download speeds are different. We point out some problems in these environment of the ALM and show their solutions at two stages. First, we redefine the "degree parameter" which originally specifies the number of single rate streams the host can transmit, to reflect asymmetric link properties and to accommodate layered streams. Second, we develop an efficient tree construction algorithm using the renewed degree parameter, which maximizes total throughput of all hosts. Simulations using NS-2 prove improvement in throughput, delay and overhead by our proposal. Software implementation is also carried out, and subjective quality improvement is verified. © 2005 IEEE.

Content Distribution Overlays improves end-user performance by replicating Web contents on a group of geographically distributed sites interconnected over the Internet. However, with the development whereby overlay systems can manage dynamically changing files, an important issue to be resolved is consistency management, which means the cached replicas on different sites must be updated if the originals change. In this paper, based on the analytical formulation of object freshness time, web access distribution and network topology we derive a novel algorithm as follows: (1) For a given content which has been changed at its original server, only a limited number of its replicas instead of all replicas are updated. (2) After a replica has been selected for update, the latest version will be sent from an algorithm-decided site instead of from its original server. Simulation results verify that the proposed algorithm provides much better consistency management than conventional methods with the reduced update overhead and network traffic.

Content Distribution Overlays improves end-user performance by replicating Web contents on a group of geographically distributed sites interconnected over the Internet. However, with the development whereby overlay systems can manage dynamically changing files, an important issue to be resolved is consistency management, which means the cached replicas on different sites must be updated if the originals change. In this paper, based on the analytical formulation of object freshness time, web access distribution and network topology we derive a novel algorithm as follows: (1) For a given content which has been changed at its original server, only a limited number of its replicas instead of all replicas are updated. (2) After a replica has been selected for update, the latest version will be sent from an algorithm-decided site instead of from its original server. Simulation results verify that the proposed algorithm provides much better consistency management than conventional methods with the reduced update overhead and network traffic.

This paper presents a novel approach for efficient tree construction in ALM (Application Layer Multicast) using layered video coding. This system is suitable for the users in heterogeneous environment such as xDSL and CATV, where upload and download speeds are different. We point out some problems in these environment of the ALM and show their solutions at two stages. First, we redefine the "degree parameter" which originally specifies the number of single rate streams the host can transmit, to reflect asymmetric link properties and to accommodate layered streams. Second, we develop an efficient tree construction algorithm using the renewed degree parameter, which maximizes total throughput of all hosts. Simulations using NS-2 prove improvement in throughput, delay and overhead by our proposal. Software implementation is also carried out, and subjective quality improvement is verified.

TCP Reno versions are widely used in current network, however it has been actualized that their throughput deteriorates in high-speed network and wireless environments. To overcome these problems of TCP Reno versions, a number of protocols have been proposed. In these researches, friendliness with TCP Reno becomes important. TCP Westwood is an example that is based on end-to-end bandwidth estimate, and brings higher efficiency performances. However it will be shown that the friendliness with TCP Reno is deteriorated according to network situations such as the buffer size of a bottleneck link router. In this paper, we quantify the buffer size that TCP Reno and TCP Westwood perform friendly to each other, and then propose an improved version of TCP Westwood that achieves friendliness with TCP Reno by estimating the buffer size of a bottleneck link router by using bandwidth estimation technique, RCE, and by updating congestion control parameters under the constraint of throughput estimation models. We confirm effectiveness of our proposal scheme by extensive simulation experiments. Simulation results show that our proposed scheme performs friendly to TCP Reno in various network situations independently of router buffer sizes.

CDN (Content Delivery Networks) improves end-user performance by replicating web contents on a group of geographically distributed servers. However, repeatedly keeping the entire replica of the original objects into many content servers consumes too much server resource. This problem becomes more serious for the large-sized objects such as streaming media, e.g. high quality video. In this paper, we therefore propose an efficient replication method for layered video streams in CDN, which can reduce user response delays and storage costs simultaneously. Based on an analytical formulation of the cooperative replication of layers and segments of each video stream, we derive a replication algorithm which solves next three problems quantitatively. (1) How many servers should be selected to replicate a given video stream? (2) For a single video stream, how many layers and segments should be stored in a given server? (3) After selecting a group of servers for each video stream, how do we allocate the replication priority (i.e. order) to each server? Simulation results verify that the proposed algorithm efficiently resolves the above problems and provides much better performance than conventional methods.

In this paper, an efficient algorithm to look for backup parents in preparation of parent leaving is proposed for application layer multicasting whose topology is constituted in the shape of a tree from a single source node. In most conventional methods, each child node starts searching for its new parent after its parent node leaves from a multicasting tree. This reactive operation often causes long interruption period. In our proposal, each node holds its parent candidate proactively over redundant overlay trees. Proactive route maintenance leads to smooth switching to a new parent after node leaving and failure, and redundant structure of a multicasting tree avoids exhausting search of a backup parent. Computer simulations are also carried out and effectiveness of the proposed approach is verified.

With the advance of network technologies, availability and popularity of streaming media contents over the P2P (Peer-to-Peer) Networks have grown rapidly in recent years. However, how to efficiently search a requested steaming media among P2P peers is still a problem which causes a serious user delay and limited hit ratio. This paper presents an efficient search method for streaming media in P2P, which reduces user response delays and exchange overhead simultaneously. Based on an analytical formulation of both streaming media and P2P peers' characteristics, we derive a search algorithm which solves the next two problems quantitatively. (1) How to decide the number of walkers (queries) at each step of search? (2) How to decide the length of walkers (queries) at each step of search? Simulation results verify that the proposed algorithm efficiently resolves the above problems and provides much better performance than conventional methods.

近年,多くの携帯電話端末に3Dグラフィックス表示機能やJava実行環境が備わり,さらにはMPEG 4ビデオの符号化・再生機能も備えられている.本稿では,これらの機能を統合することで,より魅力的なコンテンツを作成・提供することを目的とする,携帯端末上の動画・グラフィックス統合システムに関する報告を行う.本検討の特徴として,端末に大幅な機能拡張を求めることなく,既存の機能を活用するシステム構成を提案するとともに,PC上で行ったシミュレーション結果を報告する.

近年,多くの携帯電話端末に3Dグラフィックス表示機能やJava実行環境が備わり,さらにはMPEG4ビデオの符号化・再生機能も備えられている.本稿では,これらの機能を統合することで,より魅力的なコンテンツを作成・提供することを目的とする,携帯端末上の動画・グラフィックス統合システムに関する報告を行う.本検討の特徴として,端末に大幅な機能拡張を求めることなく,既存の機能を活用するシステム構成を提案するとともに,PC上で行ったシミュレーション結果を報告する.

The efficient distribution of stored information has become a major concern ill the Internet. Since the web workload characteristics show that more than 60% of network traffic is caused by image documents. how to efficiently distribute image documents from servers to end clients is an important issue. Proxy cache is an efficient solution to reduce network traffic. 2 And it has been shown that an image caching method (Graceful Caching) based oil hierarchical coding format performs better than conventional caching schemes ill recent years. However, as the capacity of the Cache is limited, how to efficiently allocate the cache memory to achieve a minimum expected delay time is still a problem to he resolved. This paper presents all integrated caching algorithm to deal with the above problem for image databases. web browsers, proxies and other similar applications ill the Internet. By analyzing the web request distribution of the Graceful Caching, both replacing and pre-fetching algorithms are proposed. We also show that our proposal can be carried out based oil information readily available ill the proxy server; it flexibly adapts its parameters to the hit rates and access pattern of users requesting documents in the Graceful Caching. Finally we verify the performance of this algorithm by simulations.

With the advance of high-speed network technologies, availability and popularity of streaming media contents over the Internet has grown rapidly in recent years. Because of their distinct statistical properties and user viewing patterns, traditional delivery and caching schemes for normal web objects such as HTML files or images can not be efficiently applied to streaming media such as audio and video. In this paper, we therefore propose an integrated caching scheme for streaming media with segment-based caching and hierarchically distributed proxies. Firstly, each stream is divided into segments and their caching algorithms are considered to determine how to distribute the segments into different level proxies efficiently. Secondly, by introducing two kinds of segment priorities, segment replacing algorithms are proposed to determine which stream and which segments should be replaced when the cache is full. Finally, a Web-friendly caching scheme is proposed to integrate the streaming caching with the conventional caching of normal web objects. Performance of the proposed algorithms is verified by carrying out simulations.

This paper presents a novel framework for Euclidean structure recovery utilizing a scaled orthographic view and perspective views simultaneously. A scaled orthographic view is introduced in order to automatically obtain camera parameters such as camera positions, orientation, and focal length. Scaled orthographic properties enable all camera parameters to be calculated implicitly and perspective properties enable a Euclidean structure to be recovered. The method can recover a Euclidean structure with at least seven point correspondences across a scaled orthographic view and perspective views. Experimental results for both computed and natural images verify that the method recovers structure with sufficient accuracy to demonstrate potential utility. The proposed method can be applied to an interface for 3D modeling, recognition, and tracking.

This paper presents a system architecture for synthetic/natural hybrid coding toward future visual services, Scene-description capability, terminal architecture, and network architecture are discussed by taking account of recent standardization activities: MPEG, VRML, ITU-T, and IETF, A consistent strategy to integrate scene-description capability and streaming technologies is demonstrated. Experimental results are also shown, in which synthetic/natural integration is successfully carried out.

© 1992 IEEE. New parallel filter banks allowing perfect reconstruction which are categorized in coding intermediate between subband coding and predictive coding are presented. M-channel filter banks allowing perfect reconstruction are represented mathematically, and filter banks categorized in the above coding are discovered. Three types of three-channel filter banks are designed. The coding gains of these filter banks are then maximized, i.e., the filter coefficients are optimized from the viewpoint of energy compaction property. Computer simulation shows that the designed filter bank performs well.

本研究において目的とするモニタリングアーキテクチャは,情報基盤,社会基盤,人々の日常生活を守るためのソフトウエアの一群である.通常は独立に研究されていた研究分野を統合することにより,従来解決が困難であった問題を解決していくことを可能とする.本研究では,計算機システム,ネットワークシステム,実世界に関する統合的なスケーラブルモニタリングに関するシステム構築をおこなうことで将来の計算機システムのあるべき姿を検討した.

H17年度は、マルチホップ無線ネットワークにおけるマルチパスルーティングプロトコルに関して、シミュレーションよる特性評価と、Linuxを積んだノートPCへのソフトウェア実装に重点を置いた検討を進めた。そして、複数台のノートPCから構成される実際のマルチホップ無線ネットワークを構築し、学内においてビデオ信号の配信実験を行った。具体的には、以下の項目の検討を行なった。
(1)通信状況に適応して動的にパスを切替えるマルチパスルーティングのシミュレーション評価。
(2)通信状況の変化をパス毎の受信電力の値に応じて判断するマルチパスルーティングのシミュレーション評価。
(3)Linux上で動作するマルチホップ経路制御プロトコルであるKerne1AODVの実装とカスタマイズ。
(4)複数台のノートPCで構築したマルチホップ無線ネットワーク上のビデオ配信実験の評価。
H18年度は、マルチホップ無線ネットワークにおけるマルチパスルーティングプロトコルに関して、いくつかの成果発表を行なうと共に、映像圧縮方式としてH.264/AVCを用いたビデオ配信実験を行なった。
また、実装評価実験からのフィードバックとして、実装評価環境を模擬したシミュレーション評価を実施した。具体的には、以下の項目の検討を行なった。
(1)提案するマルチパスルーティングとマルチメディア配信実験応用の成果をまとめた論文発表。
(2)H17年度に使用したH.263+の替わりにH.264/AVCを用いたマルチメディア配信実験の遂行。
(3)実装実験環境の特徴を反映したマルチホップ無線ネットワークのシミュレーション評価。
また、本申請で検討を進めるアドホックネットワークの応用例としてビジュアルセンシングシステムを想定したプロトタイプ実装も開始している。

本研究では、特にディジタル映像コンテンツのインターネット上のストリーミング配信を想定し、(1)ストリームメディアの効率的なキャッシング技術、(2)ストリームメディアの効率的な配信技術、の二課題を重視した研究開発を進めた。
まず、ストリームキャッシュ技術に関しては以下の成果を得た。(H15年度)階層的符号化を施したストリームデータの分散キャッシュに関して、サーバ毎のストレージ容量を制約として課した条件下における最適分散配置方式の提案を行った。(H16年度)H15年度の発展課題として、ストリームデータの分散検索、ストリームデータの分散複製、キャッシュの一貫性管理、の三方式の提案を行った。(H17年度)モバイル、ユビキタス環境を想定したストリームデータの協調キャッシュ方式の提案を行った。それぞれの提案について国内および国際学会で発表を行い、かつ三件の論文発表を行った。さらに、現在は一件の論文投稿を準備している。
次に、ストリーム配信技術に関しては以下の成果を得た。(H15年度)IP放送時のノード障害対策として、ノード障害時の経路再構築を、短時間にかつ少ないオーバーヘッドで効率的に実現できるプロアクティブ経路管理方式の提案を行った。(H16年度)ダイヤルアップからFTTHまで多様な接続環境にある加入者の要望を満たすために、階層的符号化の導入によって多様な接続形態に柔軟に適応しうるIP放送方式の提案を行った。(H17年度)最終年度として、H15年度とH16年度の成果を結合してPC上にソフトウェア実装を行い、実ネットワーク上の評価実験を行い、その結果として提案方式の有効性を実証した。それぞれの提案について国内および国際学会で発表を行うと共に、現在は二件の投稿論文が査読中になっている。

本年度は、アドホックネットワークにおけるマルチパスルーティング方式の特性改善とソフトウェア実装、およびアドホックネットワークにおけるTCPフレンドリ制御の特性改善、に重点を置いて研究開発を進めた。
(1)アドホックネットワークにおけるマルチパスルーティング:昨年度より開発を進めているマルチパスルーティング方式に関して、IETF MANETグループでRFC化されているAODV(Ad-hoc On-demand Distance Vector)方式をベースに、主にルート選択時のメトリックの改善とシグナリングパケットの減少を目的として方式の開発を進め、シミュレーション評価によって有効性を明らかにした。また、並行してLinuxノートパソコン上へのソフトウェア実装を進め、本年度中の完成を目指して作業を進めている。実装完了時には、シミュレーション評価との条件を揃え、提案方式の実装評価としての有効性を実証する予定である。
(2)アドホックネットワークにおけるTCPフレンドリ制御の特性改善:近年、国際学会において件数が増加しているIEEE802.11のモデリングと解析評価に着目し、VoIPやビデオ配信を想定したTCPフレンドリへの応用について検討を行い、通常の輻較によるパケット廃棄と、ホストの移動に伴う経路更新によって生じるパケットの順序逆転を分類するアルゴリズムを提案し、シミュレーション評価によって提案方式の有効性を明らかにした。
成果発表としては、(1)に関しては、アルゴリズムに関して国際学会発表を行い、実装評価終了時には論文投稿を予定している。(2)に関しては、国内学会への投稿を行い、今後は国際学会投稿を行うと共に、実装評価への反映と論文化を予定している。

動画像キャッシュ技術、ならびに関連する技術について、以下の検討を行った。
(1)動画ストリーミングに適した分散・階層化キャッシュアルゴリズムの提案と評価
現状のWebキャッシュメカニズムでは、キャッシュサーバが全データをキャッシュするため、動画像のような莫大なデータになると効率が大きく低下する。そこで、動画像データをセグメントと呼ばれる小単位に分割してキャッシュを行う方式、キャッシュサーバを分散的・階層的に配置し、サーバ間の連携によってストリームキャッシュを実現する方式、等を提案した。さらに、検討結果をまとめて国際学会で発表すると共に、電子情報通信学会論文誌への投稿を行なった(採録決定)。
(2)Webフレンドリなストリームキャッシュアルゴリズムの提案と評価
上記の方式はストリーミングメディアに対しては大きな改善結果をもたらすが、その反面、既存のWebデータのキャッシュ効率を低減させる危険性がある。そこで、キャッシュ容量におけるストリーミングデータのキャッシュ量を制限し、Webキャッシュ、ストリームキャッシュそれぞれの効率を高く保つ方式(Webフレンドリなストリームキャッシュ方式)に関する検討を進めた。さらに、検討結果をまとめて国際学会で発表すると共に、(1)と合わせて電子情報通信学会論文誌への投稿を行なった(採録決定)。
(3)階層化画像データベースにおけるキャッシングアルゴリズムの提案と評価
階層的符号化を活用した画像データベースにおける階層化データの置き換え(リプレース)アルゴリズムと先読み(プリフェッチ)アルゴリズムに関する検討を行った。具体的には、嗜好度に応じて階層化データの置き換え、先読みの実行を判断するアルゴリズム、嗜好度に応じたWebアクセスモデルに基づく理論解析手法、等を提案した。さらに、検討結果をまとめて国際学会で発表すると共に、電子情報通信学会論文誌への投稿を行なった(採録決定)。
(4)ストリーミングに適したCDNサーバ間のデータ複製アルゴリズムの開発
近年大きな注目を集めているCDN(Content Delivery Network)に着目し、人気の度合いに応じたCDNサーバ(サロゲート)間のストリーミングデータの効率的な複製アルゴリズムの検討を開始し、ユーザ応答時間や各サーバにおけるデータ格納コストを低減できることを明らかにすると共に、学会発表を行なった。

前年度の研究成果を発展させ、無線環境、ならびに有線・無線の混在環境下においてインターネット接続を行うときに、効率的なマルチメディア通信を実現するための諸問題の解決に努めた。
(1)移動環境に適した受信者駆動階層化マルチキャストプロトコル
IPマルチキャストを用いたマルチメディア情報の配信制御方式として、受信者駆動階層化マルチキャスト(Receiver-driven Layered Multicast : RLM)プロトコルが知られている。本提案では、このRLMプロトコルを移動通信環境でも使用可能とするための二つの方式を提案した。
・接続エリアの異なる複数のドメインを用意し、そのドメインと各階層を対応つける方式。
・RLMプロトコルの状態遷移に、無線環境特有のビット誤りによって発生したパケット廃棄による状態遷移を付加した方式。
それぞれの方式に対してシミュレーションを行い、提案方式の有効性を明らかにした。
(2)有線・無線境界部(ゲートウェイ)における適応誤り制御技術
有線・無線の混在したインターネット接続環境を想定し、有線接続を想定したパケット廃棄対策として使用されるFECパケット方式と、無線接続を想定したビット誤り対策として使用されるFECヘッダ方式とを統合し、それぞれを有線・無線境界部のゲートウェイにおいて終端・適応制御をかけることで、FEC(誤り訂正符号)によるオーバーヘッドの影響を最低限に抑え、かつ受信端末における満足のいく再生品質を実現するための方式を提案した。合わせて、シミュレーションによって提案方式の有効性を明らかにした。また、既存のRTP/RTCPプロトコルを用いたゲートウェイの実装手段も明らかにした。
(3)高速ハンドオフ・シームレスハンドオフ技術の開発
移動しながらインターネット接続を行う場合に、特に音声や動画像の受信時に大きな品質劣化要因となるハンドオフ(基地局の切り替え)の問題に対して、ハンドオフの影響を最低限に抑えるための移動管理アルゴリズムの提案を三件行った。
・ハンドオフ時に旧ドメインと新ドメインに小規模なマルチキャストを行う方式。これによって接続の瞬断が軽減される。
・ハンドオフ時にアクセスルータと基地局の双方で階層的にバッファリング・バイキャスティングを行う方式。これによって、パケットの誤順序到着が抑えられる。
・ハンドオフ接続先の予測にGPS等の位置情報を利用して接続ドメインの再構成を行う方式。これによって、物理層レベルのアクセスルータのドメイン構成とネットワーク層レベルのドメイン構成の不一致の問題が解消される。
以上の提案方式それぞれに対してシミュレーション評価を行い、提案方式の有効性を明らかにした。

画像の画質評価尺度として、PSNRの計測には原画を用意する必要があるが、YouTubeに代表されるCGM(Consumer Generated Media)映像では原画を用意することが困難な場合が多い。そこで申請者はPSNRを変化させた多数の圧縮画像をSVM(Support Vector Machine)に学習させ、さらに顕著性マップを用いて、画像を顕著性領域と非顕著領域に分割することで推定精度の改善を図るブラインド（NR型：non-reference型）画質推定に関する検討を進めてきた。また、顕著性マップでは全体的に平坦な画像もしくは複雑な画像であっても相対的に顕著性領域の検出を試みるため、適切な分割が行われず、精度が低下することがあった。そこで、本申請ではまた、動画において各フレームを杭州貼れ別を用いて2つに分割し、領域毎にブラインドPSNR推定を行う方式の検討も行った。それによって、顕著性マップによる分割が失敗する場合でも、安定したPSNR推定が行えることを確認した。また、画像の高能率符号化方式では、低ビットレート時にブロック処理に起因したブロック歪みが圧縮画像中にしばしば発生する。圧縮画像の客観画質評価尺度としてPSNRやSSIMがよく用いられているが、これらはブロック歪みに影響された画質劣化を評価することは難しく、主観画質評価とは大きな差が発生してしまう。また、圧縮画像におけるブロック歪みの影響を評価する様々な手法が提案されているが、現時点で確立した手法があるとは言い難い。申請者は地デジ映像をキャプチャした多数のMPEG-2圧縮画像を対象に、収集した多数の主観評価結果に基づき、従来手法から求めた各種の画像特徴量とブロック歪みの主観評価結果の相関性について確認した。ただし、テストシーケンスの評価が曖昧かつMPEG-2でのみ圧縮したものであったため、検証は不十分であると考えた。そこで本申請ではまた、様々な動画像符号化により生じたブロック歪みを有する動画像に対して、ブロック歪みを検証する従来手法で求めた各種の画像特徴量により、手法の有効性確認及び性能比較を行った。

現在、圧縮画像の画質評価手段として、PSNR(Peak-to-Peak Signal-to-Noise Ratio)やSSIM(Structural SIMilarity )が使用されている。しかし、これらの評価手段では、ブロック歪みなどの主観要因に特化した画質の劣化を評価することは難しい。そこで本研究では、ブロック歪みを有する画像の評価手段として、ブロック歪みの周期性に着目したケプストラム分析法を用いた検討を行った。具体的には、ブロック歪みを有するテスト画像のケプストラム情報を分析し、現在広く普及している各種の動画像符号化方式（H.261、MPEG-1、MPEG-2、H.263、MPEG-4、H.264/AVC、H.265/HEVC）に対して各々ケプストラム解析を行い、ブロック歪み評価手法の有効性確認とコーデックの性能比較を行った。その結果として、ケプストラム分析によってブロック歪みの影響を定量評価できること、動画像符号化方式の進歩に合わせてブロック歪み量が低減していること、などを示した。また、FlickrやYouTubeなどのCGM(Consumer Generated Media)を参照した画像処理として、自然画像を用いたマーカベースAR(Augmented Reality)の検討を進めている。これは、スマートフォンなどの画像入力に対し、画像内のオブジェクトをマーカーとしてCGMデータベース内の画像検索を行い、入力画像のスケールと方向を推定し、入力画像上にCGをオーバーレイ表示する方式である。本年度は、参照する画像のひずみの影響の検証を行い、画像のボケや反射の影響に関する評価実験を行った。その結果、画像のひずみ量に応じて検出可能な特徴点数が減少し、適切なAR表示が困難になること、より多数の画像群を参照することで、ひずみの存在する環境下でも正しい画像を選択できること、などを示した。今後は画質推定と画像処理とを組み合わせ、予測された歪み量に応じて適切な処理を選択する適応処理に関する検討を進めていく予定である。

本研究では、圧縮画像の客観画質評価手法として、原画像を必要とせずに圧縮ストリームの符号化情報をもとに画質推定を行うNR(Non-Reference)型画質推定に着目し、以下の検討を行った。具体的な画質評価尺度としては、古くから広く用いられているPSNR (Peak-to-peak SNR)と、近年PSNRよりも主観評価に近い客観尺度として注目されているSSIM (Structural Similarity) を対象にしている。(1) NR型PSNR推定方式のH.264/AVC圧縮動画像への適用(2) NR型PSNR推定方式のJPEG圧縮画像への適用(3) NR型SSIM推定方式の定式化とJPEG圧縮画像への適用(1)では、既存のMPEG-2圧縮を対象とするNR型PSNR推定方式を、H.264/AVC圧縮に拡張し、実画像を用いた評価実験を行った。その結果、他の競合方式に比べて推定のばらつきはあるものの、NR型PSNR推定をH.264/AVCに適用するにあたっての、ある程度の推定精度の見通しを得た。(2)では、(3)の実験の前段階として、NR型PSNR推定方式のJPEG圧縮への適用を試みた。方式自体はMPEG-2方式の簡略化であり、実画像を用いた評価実験としても、良好な推定結果を得た。(3)では、従来にはない新規提案として、NR型SSIM推定の推定アルゴリズムの定式化と、実際のJPEG画像を用いた評価実験を行った。その結果として、正確な統計量を必要とするSSIMの推定アルゴリズムと、画像の局所性に依存して値が変動するSSIMの実測値の間にはトレードオフの関係（推定領域が大き過ぎると局所性が反映されず、小さ過ぎると統計量がばらつく問題）があることを明らかにした。今後の課題としては、Flickrのような大規模画像データベースを対象にNR型PSNR推定を適用することと、NR型SSIM推定の精度改善を試みる。

本研究では、圧縮画像の画質改善技術として、以下の検討を行った。(1) 圧縮画像の複数イントラフレームを重ね合せることによる画質改善(2) 圧縮画像の圧縮ストリームからのNR型画質推定(3) 圧縮画像の周波数補間による高解像度化(1)では、量子化誤差の統計的独立性を利用して、復号側における動き検出と位置合わせを行い、複数フレームを重ね合せることで、量子化誤差を低減する方式を示し、効果を実証した。特に複数個の動き検出アルゴリズムを提案・比較し、有効性確認を行った。(2)では、動画像の圧縮ストリームから変換係数と量子化ステップサイズの情報を取り出し、原画像を参照することなく、統計処理によって圧縮画像の画質推定を行うNR(Non-Reference)型画質推定に関する検討を行った。特に対象をH.264/AVCとし、その圧縮アルゴリズムに適合した画質推定技術の開発と評価を行った。(3)では、離散コサイン変換ベースの画像拡大手法において、拡張領域のパディング法に工夫を加えることで高解像度画像の画質改善を実現する方式について検討を行った。

映像センサーネットワークにおける映像処理技術、映像配信技術、ならびにその応用技術に関する検討を行った。映像処理技術に関しては、互いに独立に符号化を行った複数枚の圧縮画像を、互いに位置合わせを行った上で重み加算することにより高画質化を図る（量子化誤差を減らす）方式を提案し、各種の成果発表を行った。この方式は、分散センサー環境において、複数のカメラが同時に撮影した圧縮画像から作成したパノラマ画像の高精細化や、一台のカメラが撮影した圧縮動画像の高画質化に適用可能である。この研究成果は、信号処理の分野で著名なIEEE ICASSP 2008にも採択され、2008年4月の成果発表を予定している。映像配信技術に関しては、無線マルチホップ接続に適した輻輳制御方式の検討を行い、シミュレーションとLinux実装の併用による各種輻輳制御方式の比較検討を行ない、今後の研究開発に向けた問題点の整理を行った。また、本件とは別の研究成果として、シングルホップ接続を前提に、独自のRTT公平性とスループットの高効率性を実現する輻輳制御方式の成果発表を行っており、現在は、その方式のマルチホップ拡張の検討を進めている。さらには以上の基盤技術の応用として、無線受信強度を用いた位置検出技術を併用した映像センサーアプリケーション、ならびに、無線受信強度を用いた混雑推定技術を併用した映像センサーアプリケーションを検討し、それぞれのプロトタイプ試作を行った。共にWeb2.0的な手法も取り入れ、各種センサーが取得した情報をサーバ上に集約し、その集約データに対して、リモートにユーザがアクセス可能なWebインタフェースを実装した。

　インターネットではネットワークのふくそうが発生し、パケットの転送が遅れたり、パケット廃棄が発生することは避けられない。これはリアルタイム指向のマルチメディア通信では大きな品質劣化の原因となる。そこで、本研究では、これらの影響を回避・低減し、インターネット上で効率的なマルチメディア通信を実現するための帯域制御技術、ふくそう制御技術について検討を行った。従来方式としては、パケット廃棄が起きてから転送レートを下げる方式、ラウンドトリップ遅延を測定しながら許容値を超えたらレートを下げる方式、などが知られている。しかし、前者はレートの変動が激しすぎて品質が安定せず、後者はラウンドトリップの許容値を自動的に算出できないために任意のネットワークに適用できない、という問題があった。本研究では、特に後者の問題を解決するために、ラウンドトリップ遅延を判断基準として用いながら動的に転送レートを更新する、任意のネットワークに自由に適用できるアルゴリズムを検討した。具体的には、二通りの方法を検討した。第一の方式では、ネットワーク全体をバッファとみなし、その容量を計測可能なラウンドトリップ遅延と受信レートから動的に更新し、それに従って転送レートを制御する。第二の方式では、ラウンドトリップ遅延を継続的に測定し、それから求まる平均ラウンドトリップ遅延が一定の範囲に収まるように、転送レートをフィードバック制御する。さらに、これらについて実際にインターネット上における転送実験を行い、それぞれの方式の有効性を確認した。