基于子带能量累积变化的语音端点检测

噪声环境下的语音端点检测在稳健语音识别中占有十分重要的地位。根据噪音和语音子带能量的累积分布变化,提出一种新的语音信号端点检测算法。通过计算各帧的子带能量变化程度,并以此设定门限进行语音端点的检测。实验表明,与一些传统的端点检测算法比较,该算法在速度和抗噪声能力上都有所增强,适合低信噪比下的语音端点检测。

Speech Endpoint Detection Based on Sub-band Energy Distribution

Accurate speech endpoint detection in adverse environments is very important for robust speech recognition. Based on the sub-band energy distributions and cumulative variations between noise and speech,a novel method for speech endpoint detection is proposed in this paper. For detecting the voice activity,a threshold is determined by computing sub-band energy change level of each frame. As compared with some traditional endpoint detection methods,this method could achieve certain improvements on detection s...