Reserarch on Multimedia Mediation Mechanism for Relization of Human-oriented Information Environments

实现人性化信息环境的多媒体中介机制研究

基本信息

批准号：
09NP1401
负责人：
SAKAUCHI Masao
金额：
--
依托单位：
The University of Tokyo
依托单位国家：
日本
项目类别：
Grant-in-Aid for Creative Basic Research
财政年份：
1997
资助国家：
日本
起止时间：
1997 至 2001
项目状态：
已结题

来源：
https://kaken.nii.ac.jp/grant/KAKENHI-PROJECT-09NP1401/
关键词：
Multimedia System Stream Type Environments Digital Broadcast Internet Multimedia Search Computer Vision Realworld Type Environment Video Information Processing マルチメディアシステム

项目摘要

In this research project, following major results contributing to new information services and systems have been obtained in three promising multimedia environments ; (1) Stream-type environments including digital broadcasting, (2) Realworld-type environments including ITS and, (3) Network type environments including WWW.(1)-1. For stream-type multimedia information, a new framework of automated video recognition/utilization, and powerful video recognition systems based on novel multimodal processing have been proposed.(1)-2. Mediation functions such as video retrieval, data collaboration, authoring etc., and several novel application systems with actual broadcasting programs have been developed.(1)-3. High performance computer architectures suitable for video and huge data processing have been proposed and implemented.(2)-1. As a primary function for event discovery for Realworld-type video data, an automated recognition system of traffic accidents and other events has been successfully developed in actual test site of " Kanda-surugashita crossing".(2)-2. New data restore methods for precise geometry and luminescence distribution of Realworld objects have been proposed and applied to outdoor cultural assets such as the Kamakura Great Buddha.(2)-3. Urban building identification methods and other realworld application systems including ITS have been developed.(3)-1. For network-type multimedia environments, efficient search engines and data mining methods for WWW data space have been developed.(3)-2. Image and audio retrieval systems in Internet space and other novel applications have been developed.Above results have been published by 1,120 technical papers and 90 mass-media articles etc.

在这个研究项目中，在三个有前景的多媒体环境中获得了以下对新信息服务和系统做出贡献的主要成果： (1)流型环境，包括数字广播，(2)现实世界型环境，包括ITS，以及，(3)网络型环境，包括WWW。(1)-1。针对流式多媒体信息，提出了一种新的自动视频识别/利用框架，以及基于新型多模态处理的强大视频识别系统。(1)-2．视频检索、数据协作、创作等中介功能以及一些具有实际广播节目的新颖应用系统已经开发出来。(1)-3．适合视频和大数据处理的高性能计算机体系结构已经被提出并实现。(2)-1．作为真实世界型视频数据的事件发现的主要功能，交通事故等事件的自动识别系统已在“神田骏下十字路口”的实际测试现场成功开发。(2)-2．提出了现实世界物体的精确几何和发光分布的新数据恢复方法，并将其应用于镰仓大佛等户外文化资产。(2)-3．城市建筑识别方法和包括ITS在内的其他实际应用系统已经开发出来。(3)-1．针对网络型多媒体环境，开发了针对WWW数据空间的高效搜索引擎和数据挖掘方法。(3)-2．开发了互联网空间图像和音频检索系统以及其他新颖的应用。以上成果已发表在1,120篇技术论文和90篇大众媒体文章等中。