摘要
采用一种基于自组织特征映射（ＳＯＭ）神经网络和隐马尔柯夫模型（ＨＭＭ）法结合的语音识别方法．根据语音信号的相关性，提出了一种自适应局部搜索算法．结果表明，该算法相对于全搜索算法而言。
A new scheme of speech recognition, which is based on the SOM/HMM principles, has been provided in the paper. An adaptive local searching algorithm is given accor ding to the correlations of speech signals. Experimental results show that the new scheme has smaller computation but has no recognition rate impairments as compared with the global searching algorithm.
出处
《上海交通大学学报》
EI
CAS
CSCD
北大核心
1996年第7期130-133,共4页
Journal of Shanghai Jiaotong University
基金
国家攀登计划认知科学(神经网络)重大关键项目
