Speech Enhancement via Bayesian Multi-solution Shrinker

摘要：

　　To effectively extract inherent information from measured speech signals, it is important to preprocess data to reduce noise.In this letter, we propose an algorithm---Bayesian multi-solution shrinker (BMS) for speech enhancement.The basic idea of BMS is to utilize empirical Bayesian model in the wavelet coefficients shrinkage step.Using speech data and calculating signal-to-noise ratio (SNR) and segmental signal-to-noise ratio (SSNR), we show that shown that the proposed approach outperforms the benchmark methods based on log-spectral amplitude (LSA), spectral subtraction and Stein”s Unbiased Risk Estimate (SURE) wavelet denoising, respectively.

关键词： Speech enhancement denoise Bayesian multi-solution shrinker

作者: Zongbo Xie Jiuchao Feng

作者单位: School of Electronic and Information Engineering,South China University of Technology,Guangzhou,510641

会议类型: 国内会议

会议名称: 2013年第四届全国通信新理论与新技术学术大会CTC2013

会议地点: 北京

会议语种:英文

页码: 89-94

在线出版日期: 2014-01-01（万方平台首次上网日期，不代表论文的发表时间）