• 主题
高级搜索  >
历史搜索 清空历史
首页> 外文会议>International conference on communications and cyber physical engineering >An Automatic Real Time Speech-Speaker Recognition System: A Real Time Approach
【24h】

An Automatic Real Time Speech-Speaker Recognition System: A Real Time Approach

机译：自动实时语音识别系统：一种实时方法

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

摘要

In this paper, an automatic speech-speaker recognition system is implemented in real time noisy environment. The database creation with personalized voice in noisy environment is done with microphone arrangement. Various techniques in preprocessing step can be used to remove noise from sampled voice signal. Mel Frequency Cepstral Coefficient (MFCC) technique is used to extract Mel Cepstral Coefficients from each speech sample and thus database is created during training phase. For testing purpose, each input sampled speech signal is mapped with stored database using Vector Quantization (VQ) and Dynamic Time Warping (DTW) techniques. Output of mapped VQ is Speaker Recognition and output of mapped DTW is Speech Recognition. Using single sampled voice, real time Speech and Speaker can be recognized. This system is very useful for various applications such as Forensic, Banking where security is at highest priority.
机译：本文在实时噪声环境下实现了语音自动识别系统。在嘈杂的环境中使用个性化语音创建数据库是通过麦克风布置完成的。预处理步骤中的各种技术可用于从采样的语音信号中去除噪声。梅尔频率倒谱系数（MFCC）技术用于从每个语音样本中提取梅尔倒谱系数，从而在训练阶段创建数据库。为了进行测试，使用矢量量化（VQ）和动态时间规整（DTW）技术将每个输入的采样语音信号与存储的数据库进行映射。映射的VQ的输出是语音识别，映射的DTW的输出是语音识别。使用单个采样语音，可以识别实时语音和说话者。该系统对于各种应用（例如取证，银行业务）非常有用，在这些应用中，安全性是最高优先级。

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