首页> 外国专利> METHOD FOR SPEECH RECOGNITION BASED ON SPEAKER AND ENVIRONMENT ADAPTATION IN MOBILE DEVICES

METHOD FOR SPEECH RECOGNITION BASED ON SPEAKER AND ENVIRONMENT ADAPTATION IN MOBILE DEVICES

机译:基于说话人和环境适应的移动设备语音识别方法

摘要

This invention relates to a speech recognition method for a small mobile terminal. Speech recognition method of the mobile terminal is small, when the small mobile terminal with a built-in small-sized speaker speech recognizer detects the environmental change or receives a speaker's speech collected or determined from the noise signal of the environment and converted to a speech signal the speech recognition mijeokeung transmitted to the server, the voice recognition server small mobile terminal by sending to calculate the model parameters adapted to the changed environment by performing a speaker or a speaker or environment adaptation to the device operation is to ensure the optimal performance of the speech recognition by using the model parameters is. ; In this case, the reliability can be improved quickly and accurately by performing the operation of the terminal or the speech recognition performance without hardware performance constraints by having charge of speaker adaptation or environment adaptation for calculating the speech recognition server is there.
机译:本发明涉及一种用于小型移动终端的语音识别方法。当具有内置的小型扬声器语音识别器的小型移动终端检测到环境变化或接收从环境的噪声信号收集或确定并转换为语音的扬声器语音时,移动终端的语音识别方法很小。通过将语音识别信号发送到服务器,语音识别服务器通过向小型移动终端发送声音,以通过执行扬声器或扬声器或环境适应设备的操作来计算适应变化的环境的模型参数,以确保最佳性能使用模型参数进行语音识别是。 ;在这种情况下,通过在那里负责计算语音识别服务器的说话者自适应或环境自适应,可以通过执行终端的操作或语音识别性能而没有硬件性能约束来快速而准确地提高可靠性。

著录项

  • 公开/公告号KR100676084B1

    专利类型

  • 公开/公告日2007-01-31

    原文格式PDF

  • 申请/专利权人

    申请/专利号KR20050003877

  • 发明设计人 지미경;김회린;

    申请日2005-01-14

  • 分类号H04Q7/24;

  • 国家 KR

  • 入库时间 2022-08-21 20:33:04

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