Endpoint detection refers to digital processing techniques to find paragraphs in a speech (phonemes, syllables, morphemes, words, etc) the location of the start point and end point. Speech segment start-stop endpoints detection is a voice analysis, an essential link in the speech synthesis and speech recognition. Traditional endpoint detection method is obtained from the WAV file voice sampling, be framed and parameter calculation of short-time energy and zero-crossing rate and activity detection. This way of working is known as offline processing methods and the realization of real time processing of the speech signal, speech signal analysis has limitations. Through the development of ActiveX controls in this article,




File list:
F2_6764.doc