首页> 外国专利> METHOD FOR DETECTING VOICE OF PORTABLE PHONE AT SERIOUS NOISE ENVIRONMENT AND DEVICE THEREOF

METHOD FOR DETECTING VOICE OF PORTABLE PHONE AT SERIOUS NOISE ENVIRONMENT AND DEVICE THEREOF

机译:在严重噪声环境中检测便携式电话语音的方法及其装置

摘要

PURPOSE: Provided is a method for detecting a voice using method for generating an analogous signal and device thereof which a voice detection function can be adapted into a prior portable phone by using a software only without addition of a hardware. CONSTITUTION: A method for detecting a voice includes (a) through (g) steps. The method extracts a voice by using a voice characteristic parameter generated from a voice vocoder. (a) step is to extract the voice characteristic parameter at the vocoder. (b) step is to generate an analogous signal by using the voice characteristic parameter. (c) step is to generate a gain of a frame analogous signal generated by summing the absolute value of the analogous signal. (d) step is to calculate a duplication average of the gain of the analogous signal by time. (e) step is to determine a voice section of a current frame by using the result of the duplication average of the gain of the analogous signal and the result of a prior frames. (f) step is to compensate a start position and a last position of the voice section output from (e) step by using the gain information. (g) step is to recompensate by using a spectrum information.
机译:目的:提供一种用于检测语音的方法,该方法用于产生模拟信号及其设备,其中仅通过使用软件而无需添加硬件即可将语音检测功能适配于现有便携式电话。构成:一种检测声音的方法,包括(a)至(g)步骤。该方法通过使用从语音声码器生成的语音特征参数来提取语音。 (a)步骤是在声码器处提取语音特征参数。 (b)步骤是通过使用语音特征参数来产生类似信号。 (c)步骤是产生通过对模拟信号的绝对值求和而产生的帧模拟信号的增益。 (d)步骤是按时间计算相似信号的增益的重复平均值。 (e)步骤是通过使用类似信号的增益的重复平均值和先前帧的结果来确定当前帧的语音部分。 (f)步骤是通过使用增益信息补偿从(e)步骤输出的语音部分的开始位置和最后位置。 (g)步骤是通过使用频谱信息进行补偿。

著录项

相似文献

  • 专利
  • 外文文献
  • 中文文献
获取专利

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