Hybrid HMM & SVM model and Its Application in Robust Speech Recognition
主 题: Hybrid HMM & SVM model and Its Application in Robust Speech Recognition
报告人: 刘敬伟 教授 (北京航空航天大学数学系)
时 间: 2008-09-24 下午 4:00 - 5:00
地 点: 理科一号楼 1303
HMM is an ad-hoc model in speech recognition, and SVM is a popular statistical model in pattern recognition. In this talk, the parellel combination modelling of these two models to achieve high recognition rate in robust speech recognition are discribed, where GMM model in HMM is combined with one-against-one SVM and one-against-all SVM respectively, and some other hybrid HMM/SVM models are also summurized.
Tea Time: 2007年9月24日15:45~16:00
地 点: 北京大学理科一号楼1384E