Hybrid Model of Continuous Hidden Markov Model and Multi-Layer Perceptron in Speech Recognition

摘要：

In order to overcome shortcomings of basic hidden markov model(HMM),a hybrid model of multi-layer perceptron (MLP)and continuous hidden markov model (CHMM) is presented which bases on basic HMM. In this hybrid mode, MLP calculates each states output probability instead of CHMM.The main purpose of this model is to improve the recognition ratio of CHMM by means of the strong of MLPs nonlinear predictive capability.Speaker independent mandarin digit speech recognition which based on the hybrid models is realized.Experimental results show that the hybrid model is efficiency and has higher recognition ratio than basic CHMM.

关键词： mandarin digit speech recognition continuous hidden markov model multi-layer perceptwn

作者: Peiling Zhang Hui Li

作者单位: College of Electrical Engineering and Automation,Henan Polytechnic University Jiaozuo 454003, China

会议类型: 国际会议

会议名称: 2009 Second International Conference on Intelligent Computation Technology and Automation(2009 第二届IEEE智能计算与自动化国际会议 ICICTA 2009)

会议地点: 长沙

会议语种:英文

页码: 1014-1017

在线出版日期: 2009-10-10（万方平台首次上网日期，不代表论文的发表时间）