| 基于 UKF滤波算法的语音增强技术 |
| |
| 引用本文： | 胡沙沙,周群彪,吕学斌,陈正茂.基于 UKF滤波算法的语音增强技术[J].四川大学学报(自然科学版),2006,43(5):996-1000. |
| |
| 作者姓名： | 胡沙沙 周群彪 吕学斌 陈正茂 |
| |
| 作者单位： | 四川大学计算机学院,成都,610064 |
| |
| 摘 要： | 基于卡尔曼滤波算法的语音增强方法在实际应用中得到了比较广泛的应用,但是这种方法要求模型必须是线性的,在实际环境中语音模型一般为非线性,为了解决这一问题,本文提出了基于UKF(Unscented Kalman Filter)滤波的语音增强算法,较好地解决了上述的问题.仿真实验结果表明,该方法在较好地消除背景噪声的同时,能保持较好地语音可懂度.
|
| 关 键 词： | 语音增强 UKF AR模型 |
| 文章编号： | 0490-6756（2006）05-0996-05 |
| 收稿时间： | 2006-01-12 |
| 修稿时间： | 2006-01-12 |
A Speech Enhancement Method Based on Unsented Kalman Filter Algorithm |
| |
| HU Sha-sha,ZHOU Qun-biao,LV Xue-bin,CHEN Zheng-mao.A Speech Enhancement Method Based on Unsented Kalman Filter Algorithm[J].Journal of Sichuan University (Natural Science Edition),2006,43(5):996-1000. |
| |
| Authors: | HU Sha-sha ZHOU Qun-biao LV Xue-bin CHEN Zheng-mao |
| |
| Institution: | College of Computer Science, Sichuan University, Chengdu 610064, China |
| |
| Abstract: | The method for enhancing speech, based on Kalman algorithm get extensive application in practical application. But the method require the model must be satisfy to be linear. The speech model is generally nonlinear in the real environment. For solving above-mentioned problem, this paper describes a speech enhancement method based on Unsented Kalman Filter algorithm. The simulation result shows its excellent performance in background noise suppression while keeping very low audible distortion. |
| |
| Keywords: | UKF |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
| 点击此处可从《四川大学学报(自然科学版)》浏览原始摘要信息 |
| 点击此处可从《四川大学学报(自然科学版)》下载免费的PDF全文 |
|
