| 听觉特性和语谱特性在说话人识别中的应用 |
| |
| 引用本文： | 王红玲,张飞云.听觉特性和语谱特性在说话人识别中的应用[J].电声技术,2009,33(6):57-59. |
| |
| 作者姓名： | 王红玲 张飞云 |
| |
| 作者单位： | 许昌学院,河南,许昌,461000 |
| |
| 基金项目： | 河南省教育厅自然科学研究资助计划项目 |
| |
| 摘 要： | 大多数说话人识别系统当由实验室走向实际应用时，环境噪声的存在会造成其识别性能下降。为了提高噪声环境下说话人识别系统的识别性能，将基于听觉特性和语谱特性的语音增强技术作为预处理器，首先对语音信号进行降噪处理，提高输入信号的信噪比。实验证明，经过降噪处理的语音信号送入说话人识别系统，提高了系统的识别性能。
|
| 关 键 词： | 说话人识别 后验Bark信噪比 临界频带 临界频带带宽 |
Speech Enhancement Method of Auditory and Speech Spectrum Characteristics Applied in Speaker Identification |
| |
| WANG Hong-ling,ZHANG Fei-yun.Speech Enhancement Method of Auditory and Speech Spectrum Characteristics Applied in Speaker Identification[J].Audio Engineering,2009,33(6):57-59. |
| |
| Authors: | WANG Hong-ling ZHANG Fei-yun |
| |
| Affiliation: | Xuchang College;Xuchang Henan 461000;China |
| |
| Abstract: | Speaker identification is critical to the highly developed information society. In real environments,noise has a detrimental effect on speech signal,and degrades the performance of speaker identification system. Speech enhancement method based on the auditory and speech spectrum characteristics is used to reduce the white noise in the front-end. Experimental results show that the performance of speaker identification system is improved with the processed speech signal in the noise environments. |
| |
| Keywords: | speaker identification posteriori Bark SNR critical band width of critical band |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
