Wataru Kameyama

Hierarchical coding scheme of video signal with scalability and compatibility

Abstract A video coding scheme that enables hierarchical video signal processing in MPEG II is proposed in this paper. Dividing each input picture into hierarchical layers, a proposed prediction is applied into each layer. The proposed coding scheme is to be characterized on the following points: (1) Non-interlaced pictures for each layer picture, even for the highest resolution layer. (2) Hierarchical motion estimation to provide precise motion vectors in less process time, to each layer of pictures. (3) Hierarchical prediction and layer pictures to provide various size of video signal for display, and to be compatible with MPEG I. (4) Up-sampling and down-sampling processes without phase distortion between neighbor layers. Other functions are almost the same as MPEG I. As the experimental result, the quality of decoded pictures highly depends on the quality of non-interlaced pictures that are provided from the pre-process.

Khmer POS Tagger: A Transformation-based Approach with Hybrid Unknown Word Handling

This paper presents an initiative research on Khmer part-of-speech tagger. We propose some modifications on applying rule algorithm of the transformation-based approach to adapt to Khmer language which is morphologically and syntactically different from the English language. Furthermore, to overcome the limited coverage of the rule-based approach in handling unknown words, we propose a hybrid approach to combine the rule-based and trigram models. Although training on a very small corpus, both proposed approaches achieve higher accuracy than the conventional methods. The tagger achieves 95.27% on training data and 91.96% on test data which includes 9% of unknown words.This paper investigates the problem of the similarity measure between very short text of sentence length, which can be used in a variety of applications that involve text knowledge representation and discovery. A novel method to calculate sentence similarity is proposed, which takes into account the semantic information, word order and the contribution of different parts of speech in a sentence. The experiment result on the set of selected sentence pairs shows that our method greatly outperforms other previously reported methods.

A proposal on open DRM system coping with both benefits of rights-holders and users

The paper proposes a rights management processing system which is independent from any usage environments and enables flexibility, reusing, and redistribution of contents with the rightful consent between rights-holders and users. An implementation is also shown to evaluate both the secureness and flexibility of the proposed system, as well as a comparison with other existing systems.

Ontology-based information extraction and information retrieval in health care domain

Ontology-based information extraction is considered as an effective method to improve the performance of information extraction (IE) systems. In order to build a better IE system using ontology-based technique, the two challenges should be most taken into account are semantic elements extraction, and ontology enhancement. Focusing on the two key points above, we propose an ontology-based information extraction system and its application information retrieval system in the health care domain. In ontology-based information extraction system, we propose a non part-of-speech based method for the semantic elements extraction and the ontology enhancement to overcome the above challenges. Experiments and Results show that all functions work well in the whole system. Many kinds of semantic elements can be extracted by the proposed extraction functions regardless parts-of-speech of words. High ratio of new semantic elements is detected for the ontology and database enrichment that yield very good results in information retrieval. The system is implemented in Vietnamese.

Proactive content caching utilizing transportation systems and its evaluation by field experiment

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.

Design of motion compensation filters of frequency scalable coding-drift reduction

This paper investigates the cause of blurring called drift, which is observed in inter-frame predicted and down scaled pictures with a scalable video coding system, from a multirate stand point of view and proposes a new motion compensation method in the decoder for drift reduction.<<ETX>>

A proposal of multimedia document architecture and video document architecture

The concepts of MDA, multimedia document architecture, and VDA, video document architecture, are proposed. Harmonization with existing multimedia architectures and integration of media are described, introducing a multimedia space. A multimedia meta-structure is centered in this multimedia space. A projection of the meta-structure to an axis or a plane of multimedia space makes a mapping to a specific media architecture. Each axis is considered as a multimedia element. As a candidate for a multimedia element in multimedia space, VDA is also proposed in order to edit video information. The time-domain structure and the scenery structure for it are introduced. The video processing model is also described.<<ETX>>

Proactive Content Caching for Mobile Video Utilizing Transportation Systems and Evaluation Through Field Experiments

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.

Holistic Feature Extraction for Automatic Image Annotation

Automating the annotation process of digital images is a crucial step towards efficient and effective management of this increasingly high volume of content. It is, nevertheless, an extremely challenging task for the research community. One of the main bottle necks is the lack of integrity and diversity of features. We solve this problem by proposing to utilize 43 image features that cover the holistic content of the image from global to subject, background, and scene. In our approach, saliency regions and background are separated without prior knowledge. Each of them together with the whole image is treated independently for feature extraction. Extensive experiments were designed to show the efficiency and effectiveness of our approach. We chose two publicly available datasets manually annotated and with the diverse nature of images for our experiments, namely, the Corel5k and ESP Game datasets. They contain 5,000 images with 260 keywords and 20,770 images with 268 keywords, respectively. Through empirical experiments, it is confirmed that by using our features with the state-of-the-art technique, we achieve superior performance in many metrics, particularly in auto-annotation.

A Proposal on Video Editing System Coping with Rights Inheritance Management for Secondary Content

This paper presents a video editing system based on the proposed rights inheritance framework for generating a secondary content. We propose rights inheritance content editor (RICE) module that provides the editing control with digital rights management (DRM) according to the granted rights during the authoring process, and manages inherited rights generation for derived contents. With our proposed rights inheritance framework and RICE, a secondary content can be generated respecting the rights of original contents.