Mei Kodama

Scalable video transcoding method with spatial updatable scalability

A scalable bitstream syntax currently is used to realize the scalable video requirements. Compared with non-scalable video decoder, a much more complex video decoder is need to decode scalable bitstreams. However, it is considered that a video decoder should implement as simple as possible to play some quality video which users want. From this view point, a scalable video coding scheme by undatable bitstream is proposed to realize the scalability without scalable video decoder. Hence, as both of higher and lower layer bitstream are consequently one layer bitstream, they can be decoded by a simple one layer video decoder. Moreover, video transcoding methods are required for high speed. Hence, we use scalability technology to transcode video data. In this paper, we pay attention to spatial resolution. We propose spatial updatable scalability and explain this algorithm. Finally, our proposed method is evaluated from the view point of the coding efficiency.

A study on the hybrid image coding and structuring methods for multimedia scalability packages communication services

Multimedia communication services based on standard information, such as MPEG-1, MPEG-2, H.261 and H.263, have been demonstrated using multimedia scalability packages, which can provide multimedia information with user functionality and availability. We propose a video information architecture coding method for these packages, called MSP (Multimedia Scalability Packages). The MSP data architecture is defined using hierarchical architectures. We consider in particular temporal feature information. The temporal scalability concept is described by both MSP data and scalability content data. As coding schemes, we consider video hybrid scalability using spatial and temporal scalability and show some scalability profiles. Finally, we explain the new video temporal function by MSP data description.

A Video Content's Creating Method using Keywords in E-learning System

Recently, network has been fast every year, so video contents services have been spreading in world wide. Especially, as the applications using video, they began to pay attention to education fields. However, there are some problems to build a market, such as, the cost of producing the video contents, the plan as learning tools and the integrated system for contents utilization, and so on. It is the important that users can get the authorization of the copyright of using them and they can use at a low price as well as possible. Then, we pay attention to the person who makes them, and the way how to make. In this paper, we propose the contents making methods based on the keywords, which a teacher defines in advance, using the part of a lecture video in a university. That ldrsldr to say, the teacher is content ldrsldr producer, and he has the copyright of it, according to various educational purposes, the authorization organization is built up. I formulate a hypothesis that the organization establishes in a university from producing learning contents to making finished digital data. The merit is that the time when they need to get the permission of using becomes short and users can also use easy. We indicate the advantage of the cost of producing contents and building the system from the view point of the production results of keyword video contents. We try to make the system in practice based on the above contents, and experiment the e-learning system over network. Of course, we set up the environment of contents delivery system in cooperation with network provider. By the network experiments, we show the efficient method to produce the worthy video contents in the e-learning system.

Consideration of transcoding using updatable scalability for selective quality video content delivery method

In this paper, we propose a video access model with guaranteed bandwidth by considering data transmission and browsing. Moreover, we propose a video content delivery method with scalable transcoding, when users use multi-quality video, and evaluate the transcoding time in our video access model. From the viewpoint of transcoding time, the influence of user access time is clear in this model, comparing the proposed method with the simulcast type by simulation experiments.

Automated archive of the lectures by MPEG2 over intranetwork in Hiroshima University

On April 2002, we have established the automated electronic archive of two-way TV lecture system at Hiroshima University and archived 600 lectures in 4Mbps MPEG2 format and 1Mbps MPEG4 format. Also we have archived 50 conventional one-room lectures and special discourse taken by DV cameras. The two-way TV lecture is typical example of hi-cost and hi-quality resource, and DV shooting of conventional lectures is the counterpoint. We compare them as e-learning resource.

A video searching system using MSP data and its evaluation of matching methods

Multimedia communication services based on standard information, such as MPEG-1, MPEG-2, H.261 and H.263, had been indicated using multimedia scalability packages, which can provide multimedia information according to functionality and availability for users. This paper shows the video searching system based on exchanging packages of MSP: multimedia scalability packages. In this system, MSP data is decoded on the server side and is sent to the search engine. Input parameters for searching are described on attached information of MSP. After a server receives that, MSP searching information and content information are decoded into the matching processor. Output MSP data is made by searching system information and searching result on the server side and sending back to the user side. We indicate the effective searching features and methods of spatial and temporal domains. After considering them from the viewpoints of process time and matching numbers by experimental results, it shows that video searching is able to be realized efficiently in the temporal domain. Finally, we show the flexibility of our search system as video searching systems.

A cached video content management method with priority estimation according to the content property in VCDS-MQ with scalable data

Video content management methods had been studied based on the priority (e.g., the number of accesses and the elapsed time) in a video content distribution system (VCDS). However, these methods cannot adapt to the variable model of the practical accesses. Then, this paper proposes a cached video content management method with priority estimation according to the content property. Basically, this method linearly estimates the cache priority to adapt to the variable model. Considering the updating status of the cached contents, two kinds of contents (i.e., a regular content and a periodical updating content) are used. In particular, two content properties are not only defined from a difference in each content popularity, but also two kinds of priority estimation methods (i.e., only a linear estimation, and both the estimation and the estimation with the condition determination for the elapsed time) are designed. Here, a LFU method is used as a conventional method (CM). Next, the evaluation models are designed in 1-day model (i.e., the variable model of the number of the accesses, that of the quality and the content-usage model). CM and the proposed method were evaluated in the simulation experiments. It was found that the proposed method was superior by the evaluation results in terms of the cache efficiency.

A consideration on contents management systems using quality transition mode in video contents utilization services by on-demand

Video contents service over a network is expanded by improvement in network speed and large capacity in recent years. Meanwhile, when users use a lot of contents, they are faced with the problem of the lack of memory capacity. Therefore, the contents management methods are important for the services. Then, we had proposed the management method using cache delivery and scalable contents. In this study, after a content utilization service model using mobile and home terminal is defined, we propose a management method using the data transition process of scalable structure. This is based on elapsed time and its content quality. We consider the efficiency in this model from a viewpoint of single quality and the quality of two.

A study on new scalabe video coding for quality control

Over the broadband internet, we often use video contents. In addition, not only broadcast but also IPTV are used in large display and PC now. In these cases, basically speaking, we think that they enjoy some video contents in fixed TV environments. Moreover, in this paper, we pay attention to “Video Contents Distribution System”, and once they receive some contents data, we have to consider how to manage them and to control the quality for several user terminals and to support any network environment in integrated video contents delivery and distribution systems. In this paper, we consider an algorithm of scalable video coding in order to control the video quality easily according to the diversity of usage of video contents. We had proposed a video transcoding method using scalable structure based on linear process in quantized domain. Although it is the point that quantization parameters are used between base layer and enhancement layer, when the efficiency of coding prediction goes down, amount of coded bits also goes on increasing at the same time, you know. Its efficiency affects the control of the video quality. In this study, after we consider those situations as noted above, we propose the concept of quantization parameters description and a new modified method of scalable video coding to provide multi-quality video to users.

A study on the method to reduce data transfer in multi-quality VOD service

In the VOD service, the low resolution video data is expected for content searching, and the high resolution video data is expected for browsing. We have presented the effect of spatial video scalability encoding in cache utilization. In this study, we report comprehensive experiments on cases of various updating procedures, and the effects of differential data transferring. As a result, if a time-cycle of downloading, transcoding and browsing is too short for all clients to download contents within, the scalability encoding method can improve the quality of service, in both of data size and consumed time.