| Speech endpoint detection is an important step in the field of speech communication. In applications, the systems usually need to find out the beginning and ending point of the speech, through analysing the input signal, so that it can collect the exact speech data, cut down the amount of data and calculating as well as the time of operation. Therefore, the research on endpoint detection algorithms of speech signal is significant.How to improve performance of endpoint detection in noisy environment is an important issue in speech recognition, especially in actual noisy environments. The performance of conventional endpoint detection methods based on short-time energy and zero-crossing rate is unsatisfactory in environments of low SNR. Neural network, as a relatively new intelligent control method, is the hot research area attracting wide attention of many experts and scholars.This paper studies the present typical speech endpoint detection algorithms, and presents a new speech endpoint detection algorithm based on wavlet analysis and neural networks. It emphasis the learning algorithm of neural networks and applies it into speech endpoint detection. The result demonstrates the effectiveness of the application of neural network method in speech endpoint detection. Futhermore, it shows the flexibility and the real-time performance of the algorithm. Finally, the algorithms studied in the paper is concluded; an ongoing improvement and anticipated development are discussed. |
