AUTOMATIC SPEECH RECOGNITION USING TIED DENSITY HIDDEN MARKOV-MODELS

被引:0
|
作者
EULER, S
机构
关键词
D O I
暂无
中图分类号
TM [电工技术]; TN [电子技术、通信技术];
学科分类号
0808 ; 0809 ;
摘要
In this paper we discuss various aspects of the application of tied density Hidden Markov Models to automatic speech recognition. Using a speaker independent, isolated word recognition system as an example, different densities are compared with respect to recognition performance and computational complexity. Next, we present an approach to replace similar densities by a common representative. Finally, results are given for extending the vocabulary using the densities derived from the basic set of words.
引用
收藏
页码:274 / 279
页数:6
相关论文
共 50 条