基于子带能熵比的语音端点检测算法  

Speech Endpoint Detection Algorithm Based on Sub-band Energy-entropy-ratio

作　　者：张毅 王可佳 席兵 颜博 

机构地区：[1]重庆市信息无障碍与服务机器人工程技术研发中心,重庆400065

出　　处：《计算机科学》2017年第5期304-307,319,共5页Computer Science

基　　金：重庆市教委科学技术研究项目基金(KJ130511);重庆市科学技术委员会项目(cstc2015jcyjBX0066)资助

摘　　要：准确地识别语音端点是语音识别过程中的一个重要步骤。在低信噪比环境下,为更好地增强语音和噪声的区分度,提高语音端点检测系统的准确率,在分析了常规子带谱熵端点检测算法的基础上结合子带能量,提出了一种基于子带能熵比的语音端点检测算法。该算法将子带能量和子带谱熵的比值作为端点检测的重要参数,以此设定阈值进行语音端点的检测。实验表明,该算法快速高效,具有较高的鲁棒性,在较低的信噪比环境下能准确地进行语音端点检测。It is an important step of speech recognition process to identify accurately the speech endpoint~ Under the en- vironment with low SNR, in order to enhance the discrimination of noise better and improve the accuracy of speech end- point detection system, this paper proposesd a new type of speech endpoint detection algorithm based on sub-band ener- gy-entropy-ratio. The proposed algorithm takes the ratio of short-time sub-band energy and sub-band spectral entropy as an important parameter of endpoint detection, and sets the threshold to speech endpoint detection. Experiments show that the algorithm is fast and efficient,and it also has strong robustness and can detect the voice endpoint under lower SNR accurately.

关 键 词：端点检测 子带谱熵 子带能量 子带能熵比 信噪比 鲁棒性 

分 类 号：TP242.6[自动化与计算机技术—检测技术与自动化装置]