• 中国计算机学会会刊
  • 中国科技核心期刊
  • 中文核心期刊

计算机工程与科学

• 人工智能与数据挖掘 • 上一篇    

车载环境下的语音端点检测

涂志强,梁亚玲,杜明辉   


  1. (华南理工大学电子与信息学院,广东 广州 510641)
  • 收稿日期:2017-06-14 修回日期:2017-11-09 出版日期:2018-10-25 发布日期:2018-10-25
  • 基金资助:

    国家自然科学基金(61701181);广东省自然科学基金(2017A030325430);广州市科技计划项目(201707010070)

Voice activity detection in car environment

TU Zhiqiang,LIANG Yaling,DU Minghui   

  1. (School of Electronic and Information Engineering,South China University of Technology,Guangzhou 510641,China)
     
  • Received:2017-06-14 Revised:2017-11-09 Online:2018-10-25 Published:2018-10-25

摘要:

为了提高车载噪声环境下语音端点检测的准确性,提出了一个基于GRURNN的神经网络结构,
对带噪语音的LogMel特征序列进行处理,实现语音与噪声的分离,从而恢复出纯净语音的LogMel特征序列;在此基础上,提出一种新的特征LogMelSum,并用该特征进行端点检测。实验结果表明,在车载环境下,本文方法具有很好的端点检测性能。
 
 

关键词: 语音端点检测, 循环神经网络, 车载环境, 人声与噪音分离

Abstract:

In order to improve the accuracy of voice activity detection in car environment, we propose a neural network structure based on gated recurrent unit-recurrent neural network (GRU-RNN) to process the Log-Mel feature sequence of noisy speech and separate speech from noise. After restoring the Log-Mel sequence of clean speech, a new feature, Log-Mel-Sum, is proposed to detect voice activity. Experimental results show that the proposed method has a high accuracy of voice activity detection.
 

Key words: voice activity detection;recurrent neural network;car environment;separation of , voice and noise