会话智能代理与语音自动识别  

Conversation Agent and Automatic Speech Recognition

作　　者：冯志伟[1] 詹宏伟[1] 

机构地区：[1]杭州师范大学

出　　处：《外语学刊》2018年第1期13-23,共11页Foreign Language Research

摘　　要：本文从会话智能代理的角度，论述语音自动识别的原理和方法，分析特征提取阶段、声学建模阶段和解码阶段的基本原理，最后介绍语音识别研究的历史与现状。From the perspective of conversation agent, this paper discusses the theory and approaches of automatic speech recognition, and analyzes the basic principles of feature abstraction stage, acoustic modeling stage and decoding stage. Finally, the past and present of speech automatic recognition are also introduced.

关 键 词：智能代理 语音自动识别 特征抽取阶段 声学建模阶段 解码阶段 

分 类 号：H087[语言文字—语言学]