基于Agent的多功能感知环境研究  

DEVELOPMENT OF MULTIMODAL PERCEPTION ENVIONMENT WITH AGENT

作　　者：王延江[1] 袁保宗[2] 唐晓芳[2] 

机构地区：[1]石油大学自动化系,山东东营257061 [2]北方交通大学信息所

出　　处：《石油大学学报（自然科学版）》2001年第3期86-89,共4页Journal of the University of Petroleum,China(Edition of Natural Science)

基　　金：国家重点自然基金资助项目 (6 978930 1)

摘　　要：多功能感知是指计算机不仅可以通过键盘和鼠标 ,而且可以用手势、表情、语音等其它方式与人进行交互。给出了统一的感知Agent模型 ,提出了一个基于Agent的多感知系统结构 ,且建立了系统运行的网络环境 。Multimodal perception means that the computer can be interacted with human by using a keyboard and a mouse, and also using gestures, facial expressions and voices. This paper proposes an Agent based multi modal perception system model. Since the Agent is an autonomous, collaborative intelligent entity and has the ability of learning, thus the system is more flexible and intelligent. A standard perception Agent model is presented, and an Agent based multi modal perception system architecture is given. A network environment is built. A multi modal perception system based on speech recognition Agent and a gesture recognition Agent is introduced.

关 键 词：AGENT 多功能感知 人机交互 模式识别 语音识别 手势识别 人工智能 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]