An Improved Speech Endpoint Detection Method Based on Cepstrum Distance
Peng Wang · Communications technology · 2008
One of the causes in reducing the capability of speech recognition(SR) systems is the noisy enviroment.As an important technology in SR,the accuracy of endpoint detection determines the SR rates in some degree.In this article,an improved method based on MFCC for endpoint detection is proposed.Short time energy and zero crossing rate are intergraded as final decision threshold on the basis of traditional endpoint detection method based on MFCC.The experiments show that the high computational efficiency and good accurate detection can be obtained under the condition of low SNR.