基于选择性注意的交叉感知信息认知计算

项目名称： 基于选择性注意的交叉感知信息认知计算

项目编号： No.61332018

项目类型： 重点项目

立项/批准年度： 2014

项目学科： 计算机科学学科

项目作者： 龚怡宏

作者单位： 西安交通大学

项目金额： 300万元

中文摘要： 随着网络通信技术的迅速发展以及移动智能终端的快速普及，构建以人为中心的感知与计算网络（简称“群智感知计算”）已成为互联网的一种崭新的应用模式和发展趋势。群智感知计算理论体系的出现在实现大规模复杂社会事件感知及认知计算方面不仅带来全新的思路和手段，同时也提出了许多新的技术问题和挑战。本课题的主要研究目标是应用群智感知网络实现对大规模复杂社会事件的交叉信息融合与认知。针对大规模复杂社会事件认知任务的共性需求和典型应用特征，本课题提出三个层次的基于选择性注意的交叉信息融合与认知计算体系架构，深入研究构建这个体系架构的关键科学技术问题。应用选择性注意交叉感知信息融合与认知计算体系架构，开发热点事件关注平台，实现重大事件现场场景重建，以及复杂事件过程拼接与重现等高附加价值应用，，由此取得对热点事件整体的正确认知及把握，为决策者做出科学的判断与决策奠定基础，为群智感知计算理论体系的发展及应用做出贡献。

中文关键词： 交叉感知;选择性注意;群智;认知计算;

英文摘要： With rapid development of network communication technology and wide availability of intelligent mobile devices, building human-centered sensoring and perception network (called crowd-sourcing cognitive computing) has emerged as a new model and trend of the Internet. The situation has lead to new possibilities in dealing with large-scale complex social events, and also caused many new problems and challenges. This research focuses on applying cognitive computing network technology to cross modality information fusion and perception of large-scale complex social events. We propose a three-layer cognitive computing framework based on selective attention and cross modality perception to serve general and common requirements in perceiving various large-scale complex social events. We will emphasize on several scientific problems associated with constructing such a cognitive computing framework. We will apply such framework to implement several high value-added services, including building platforms for following hot events, reconstructing scenes of major events, reproducing spatial-temporal evolution of complex events, etc. By enabling the users to visualize the perceived event and interact with it in a cross-time, cross-space, cross-view and cross-scale manner, the research project will lead to a useful tool that can help decision-makers to obtain correct, global and timely view of the hot event, and thus assisting their decision making process, and ultimately making contributions to the development of crowd-sourcing cognitive computing theories and systems.

英文关键词： Cross Modality Perception;Selective Attention;Crowd-Sourcing;Cognitive Computing