语音识别实现多媒体检索的系统设计

A Design of Multimedia information Retrieval System based-on Speech Recognition

设计并实现利用语音识别对多媒体信息检索的系统。该系统综合运用分类法、主题法和基于内容的多媒体检索方法,并嵌入语音识别技术,以达到综合运用多媒体低级内容特征和高级语义特征来提高检索性能的目的,同时使得人机交互更为便利。

This paper designs and puts forward a system which maks multimedia information retrieval based--on speech recognition.This system is a combination of different method,an integration of classification,subject and content based multimedia retrieval techniques. With the embedded speech recognition technique,integrated usage of low--level content feature with high--level semantic feature which aims in improving the veracity. At the same time,the inter--exchange between human and computer becomes more convenient.