| 基于TVAR模型的语音增强技术 |
| |
| 引用本文： | 石鸿凌,姜琳峰,孙洪. 基于TVAR模型的语音增强技术[J]. 武汉大学学报(工学版), 2004, 37(2): 49-52 |
| |
| 作者姓名： | 石鸿凌 姜琳峰 孙洪 |
| |
| 作者单位： | 武汉大学电子信息学院,湖北,武汉,430072 |
| |
| 基金项目： | 国家自然科学基金资助项目(60072041) |
| |
| 摘 要： | 在分析语音信号的时变自回归TVAR(Time VaryingAutoregressive)模型及其模型参数的随机演化模型的基础上,基于粒子滤波器(ParticleFilter)对TVAR模型参数的序列估计,提出了一种语音增强算法.算法通过引入反射系数,快速简捷实现了模型稳定性的判断,保证了跟踪的模型的稳定性.实验结果表明,算法可以很好地跟踪非平稳信号,采用该方法处理过的语音,信噪比SNR(Signal to NoiseRatio)明显提高,听觉质量得到了很大的改善.
|
| 关 键 词： | 语音增强 时变自回归模型 粒子滤波器 |
| 文章编号： | 1671-8844(2004)02-049-04 |
| 修稿时间： | 2003-12-25 |
Speech enhancement technique based on time-varying autoregressive model |
| |
| SHI Hong-ling,JIANG Lin-feng,SUN Hong. Speech enhancement technique based on time-varying autoregressive model[J]. Engineering Journal of Wuhan University, 2004, 37(2): 49-52 |
| |
| Authors: | SHI Hong-ling JIANG Lin-feng SUN Hong |
| |
| Abstract: | By exploring the time-varying autoregressive models and its stochastically evolving parameters, and introducing reflection coefficient to ensure model strict stability, we present a time-varying autoregressive model based speech enhancement algorithm. The algorithm relies on a particle filter estimation of the TVAR model parameters. The experimental results show that after filtered by our algorithm the signal-to-noise ratio and the quality of speech have been improved obviously. |
| |
| Keywords: | speech enhancement time-varying autoregressive particle filter |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
|
