分布式语音识别系统的架构分析和具体实现  

Architecture Analysis and Realization of Distributed Speech Recognition System

作　　者：孟建庭[1] 吴及[1] 王作英[1] 

机构地区：[1]清华大学电子工程系,北京100084

出　　处：《电声技术》2004年第8期51-53,共3页Audio Engineering

摘　　要：研究了分布式语音识别系统(DSR)的架构,实现并测试了客户端/服务器架构的分布式语音识别系统。系统采用基于段长分布的隐含马尔可夫模型(DDBHMM)的识别算法,使用多服务器、多客户端的系统架构,采用负载平衡的方式分配识别资源,能够达到稳定高效的性能。This paper makes a study of the architectures of Distributed Speech Recognition(DSR) systems, realizes and evaluates a DSR system with Client/Server architectures. This DSR system uses the speech recognition algorithm based on DDBHMM (Duration Distribution Based Hidden Markov Model). The system uses multi-server and multi-client architecture and a load balance method to assign recognition tasks to different servers, and this makes the system have an excellent performance.

关 键 词：语音识别 分布式语音识别 段长分布的隐含弥可夫模型 

分 类 号：TN912.34[电子电信—通信与信息系统]