• 主题
高级搜索  >
历史搜索 清空历史
首页> 外文会议>International conference on mining intelligence and knowledge exploration >A Study on Vowel Region Detection from a Continuous Speech
【24h】

A Study on Vowel Region Detection from a Continuous Speech

机译：连续语音中元音区域检测的研究

获取原文
页面导航
  • 摘要
  • 著录项
  • 相似文献
  • 相关主题

摘要

Vowels exhibit high sonority and loudness due to varying strength of impulse-like excitations. Acoustic events such as vowel onset point (VOP) and vowel end point (VEP) can be used to detect precise vowel regions in a speech signal. In this paper, a technique is proposed to detect vowel regions based on these acoustic parameters in a continuous speech signal. Vowels possess significant energy content in the low frequency bands of speech. The initial phase of this method consists of speech signal processing using zero frequency filtering technique. Zero frequency filtered signal predominantly contains low frequency content of the speech signal as it is filtered around 0Hz. This process is followed by the extraction of dominant spectral peaks from the magnitude spectrum around glottal closure regions of the speech signal. The vowel onset points and vowel end points are obtained by convolving spectrum of zero frequency filtered signal with first order Gaussian differentiator. The performance of the proposed vowel region detection method is compared with the existing state of art methods on TIMIT database. It is reported that this method produced relatively significant improvement in vowel region detection in clean and noisy environments.
机译：由于脉冲状激励强度的变化，元音表现出较高的声音和响度。诸如元音开始点（VOP）和元音结束点（VEP）之类的声音事件可用于检测语音信号中的精确元音区域。在本文中，提出了一种基于连续语音信号中的声学参数检测元音区域的技术。元音在语音的低频带中具有显着的能量含量。该方法的初始阶段包括使用零频率滤波技术的语音信号处理。零频滤波信号主要包含语音信号的低频内容，因为它在0Hz左右被滤波。在此过程之后，从语音信号的声门闭合区域周围的幅度谱中提取主要谱峰。通过将零频率滤波信号的频谱与一阶高斯微分器进行卷积来获得元音起始点和元音端点。将所提出的元音区域检测方法的性能与TIMIT数据库中现有的现有方法进行了比较。据报道，该方法在干净和嘈杂的环境中在元音区域检测方面产生了相对显着的改进。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱：kefu@zhangqiaokeyan.com

京公网安备：11010802029741号 ICP备案号：京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有

  • 客服微信

  • 服务号