首页> 外国专利> System for encoding voice with suppression of ruìdo integrated, and process for removing ruìdo in a voice encoder

System for encoding voice with suppression of ruìdo integrated, and process for removing ruìdo in a voice encoder

机译:集成了抑制ru u00ecdo的语音编码系统和在语音编码器中删除ru u00ecdo的过程

摘要

"The system to encode voice with suppression of ruu00ecdo integrated, and,Process for removing ruu00ecdo in a voice encoder "a system to encode voice while suppressing background noise, acoustic and a process to suppress background noise acoustically in a code of VOZ are described here. The voice encoder includes a sampler that capture groups of samples in the time domain of an audio signal.A voice activity detector operatively coupled to the sampler determines the presence or absence of speech in a current group.A processor is operatively coupled to the sampler to transform the group of audio samples in the time domain into an estimate of the power spectrum of that group.A noise model adapter operatively associated with the processor updates a model of noise in the field of frequ00b3u00eancia based on estimation of power spectrum of this group theVoice activity detector indicates an absence of speech in this group.A block of filter operatively coupled to the computing model of the noise adapter and the transformed computes a spectral enhancement filter based on power spectrum estimation of P.Otu00eancia and noise model adapted.A block operatively coupled to the processor and spectral enhancement to the block of computing filter applies the filter to the current estimate of the spectrum spectral enhancement of potu00eanThe cla. A quantizador and block encoder transforms the parameters of model of voice codec, including spectral magnitudes intensified, in a group of coded bits.
机译:集成了抑制ru的语音编码系统,以及在语音编码器中去除ru的过程。一种在抑制背景噪声,声学的同时对语音进行编码的系统和在语音中抑制背景噪声的过程。此处描述了VOZ的代码。语音编码器包括采样器,该采样器在音频信号的时域中捕获多组采样。可操作地耦合到采样器的语音活动检测器确定当前组中语音的存在或不存在。将时域中的音频样本组转换为该组功率谱的估计。可操作地与处理器关联的噪声模型适配器根据功率谱的估计更新freq u00b3 u00eancia领域中的噪声模型该组中的语音活动检测器指示该组中没有语音。一个滤波器块可操作地耦合到噪声适配器的计算模型,并且变换后的滤波器基于P.Ot u00eancia和噪声的功率谱估计来计算频谱增强滤波器。可操作地耦合到处理器并且对计算滤波器的模块进行频谱增强的模块将滤波器应用于当前es锅的光谱增强光谱量化器和块编码器在一组编码比特中转换语音编解码器模型的参数,包括增强的频谱幅度。

著录项

  • 公开/公告号BR9813246A

    专利类型

  • 公开/公告日2000-10-03

    原文格式PDF

  • 申请/专利权人 ERICSSON INC;

    申请/专利号BR19989813246

  • 发明设计人 PHILLIP M. JOHNSON;LELAND S. BLOCEBAUM;

    申请日1998-12-03

  • 分类号G10L3/02;

  • 国家 BR

  • 入库时间 2022-08-22 01:57:05

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