Research On The Techniques Of Chinese Continuous Speech Boundary Detection

During the procedure of speech recognition to the Chinese continuous speech segments, there always brings the decreasing of the recognition correctness rate due to the wrong estimation for the start and the end of a piece of speech segment. In order to solve the problem, the author studies hard and in this paper puts forward a new method, which makes use of several characters to recognize the start and the end of a continuous speech segment. The total voice-segment division (VSD) process consists of two steps: the initial VSD process and the final VSD process. The initial VSD process uses two main characters, the average instantaneous energy and the average instantaneous zero crossing rate (ZCR). to make the first recognition for the start and the end, the emphasis of which is to select the appropriate value of the threshold and the length of frame. In the final VSD process, the author compares several characters and confirms the new recognition character. Kalman filter-wave parameter. At the same time the author also puts forward a new recognition character, periodic gradual change (PGC) and uses these characters to recognize the start and the end in the mini-segment. According to calculate lots of speech segments, because the author applies the new recognition characters to the recognition procedure and selects the appropriate parameter values, the author's new method improves the recognition correctness rate for the start and the end of a piece of Chinese continuous speech segment.