摘要
本文以简化的BASIC语言语句为对象实现了连续语言识别系统。该系统分为语音信号处理和语言信号处理两部份,前一部份完成音节分割、参数提取压缩及模式匹配工作后一部份分析了BASIC语言的句法结构和语义,用以单词预测和约束。这两部份交替工作,构成了一个完整的系统。
This paper presents a continuous speech recognition system in which a modified Basic language is served as a task for speech recognition. In this system we make ful use of acoustic and linguistic knowledge in the process of recognition and get a satisfactory result.
出处
《苏州大学学报(自然科学版)》
CAS
1990年第3期352-358,共7页
Journal of Soochow University(Natural Science Edition)
关键词
连续语言
音节分割
BASIC语言
continuous speech, Syllable segmentation, sentence structure, ATN syntax, semantic analysis.