| 基于谱减法的听觉模拟的语音增强 |
| |
| 引用本文： | 陶智,赵鹤鸣,龚呈卉,吴迪.基于谱减法的听觉模拟的语音增强[J].计算机工程与应用,2005,41(4):57-59,104. |
| |
| 作者姓名： | 陶智 赵鹤鸣 龚呈卉 吴迪 |
| |
| 作者单位： | 苏州大学物理学院,苏州,215006;苏州大学电子信息学院,苏州,215006 |
| |
| 基金项目： | 国家自然科学基金资助项目(编号:60172016)，苏州大学青年基金资助项目(编号:Q3108404) |
| |
| 摘 要： | 提出了一种适于低信噪比下的语音增强算法。该算法以传统的谱减法为基础,所用减参数是根据人耳听觉掩蔽效应提出的且是自适应的。对该算法进行了客观和主观测试,结果表明:相对于传统的谱减法,该算法能更好地抑制残留噪声和背景噪声,特别是对低信噪比的语音信号。
|
| 关 键 词： | 听觉系统 掩蔽效应 语音增强 谱减法 |
| 文章编号： | 1002-8331-(2005)04-0057-03 |
Speech Enhancement Based on Spectral Subtraction of Human Auditory System |
| |
| Tao Zhi,Zhao Heming,Gong Chenghui,Wu Di.Speech Enhancement Based on Spectral Subtraction of Human Auditory System[J].Computer Engineering and Applications,2005,41(4):57-59,104. |
| |
| Authors: | Tao Zhi Zhao Heming Gong Chenghui Wu Di |
| |
| Affiliation: | Tao Zhi1 Zhao Heming2 Gong Chenghui1 Wu Di2 1 |
| |
| Abstract: | A speech enhancement method for low S/N ratio is proposed.This algorithm is based on traditional spectral subtraction.Subtract parameter is got based on the masking properties of human auditory and self-adaptive.Subjective and objective tests are made on this algorithm.The results show this algorithm is very effective to reduce the residual noise and the background noise,especially for low S/N ratio speech compared to the traditional spectral subtraction. |
| |
| Keywords: | auditory system masking properties speech enhancement spectral subtraction |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
