压缩感知分组分离语音增强.pdfVIP

204 2014 ，50（24 ） Computer Engineering and Applications 计算机工程与应用 ⦾信号处理⦾ 压缩感知分组分离语音增强 宁矿凤，王景芳 NING Kuangfeng, WANG Jingfang 湖南涉外经济学院 信息科学与工程学院，长沙 410205 School of Information Science and Engineering, Hunan International Economics University, Changsha 410205, China NING Kuangfeng, WANG Jingfang. Speech enhancement based on group-separable compressed sensing. Comput- er Engineering and Applications, 2014, 50（24 ）：204-208. Abstract ：Compressed Sensing （CS ）which is a signal sparsity-based sampling method, can effectively extract the infor- mation contained in the signal. A new method is designed for noisy speech enhancement based on the grouping separation of compressed sensing. Speech sparse expression is used in discrete Fast Fourier Transform （FFT ）domain. The algorithm can implement compression measurement and denoising in noisy speech by the design of the complex domain observation matrix and soft threshold. Sparse Reconstruction by Separable Approximation, SpaRSA algorithm is used to restore the speech signal, to achieve speech enhancement. The experiments show that the denoising signal can be compressed and recon- structed for noise signal compression refactoring. The signal-to-noise ratio can be improved greatly. The background noise can be more effectively suppressed. Key words ：speech enhancement; compressed sensing; group-separable; soft threshold; denoising 摘 要：压缩感知（Compressive Sensing ，CS ）是一种基于信号稀疏性的采样方法，可以有效提取信号中所包含的信 息。提出了一种分组分离压缩感知语音增强新算法。算法利用语音在离散快速傅里叶变换（Fast Fourier Transform ， FFT ）域下的稀疏性，设计复域观测矩阵与软阈值对带噪语音进行压缩测量与去噪，通过可分组分离逼近稀疏重建 （Sparse Reconstruction by Separable Approximation ，SpaRSA ）算法恢复语音信号，实现语音增强。实验表明：该算 法对含噪信号压缩重构，信噪比幅度较大提高，能更有效地抑制背景噪声。 关键词：语音增强；压缩感知；分组分离；软阈值；去噪 文献标志码：A 中图分类号：TN912.35 doi ：10.3778/j.issn.1002-8331.1309-0040 1 引言 滤波，以及相继发展起来的子空间增强、小波变换和这 语音是非平稳、时变的信号。通过语音传递信息是