关键词: Teager能量算子, 端点检测, 语音识别, 噪声

Abstract: Speech endpoint detection is a crucial component in speech recognition system, especially in noisy environment. Its accuracy affects the computational complexity and the recognition performance of the speech recognition system. This paper proposes an endpoint detection of speech signals based on short-time TEO energy in noisy environment. It uses a three-state transition and judgment mechanism based on double thresholds, which ensures the accuracy in noisy environment and the robustness to changes in absolute levels. Compared with same traditional algorithms such as short-time energy and spectral entropy, experiment results show this algorithm has better detection capability in low signal to noise ratio environments and takes on more advantages.

Key words: Teager energy operator, endpoint detection, speech recognition, noise