| 总结: | Most of the accurate method for the speech enhancement design mainly focuses on quality and intelligibility to produce high performance level by using compression techniques. A novel speech enhancement algorithm using compressive sensing (CS) is different paradigm from compression technique with low-dimensional geometry for transmission or storage. The CS algorithm, can directly acquire compressed data signals and replace samples by more general measurements of the uniform rate digitization with signal sparsity model. Perceptual evaluation of speech quality (PESQ) is an objective evaluation of speech enhancement algorithm used to measure the enhanced speech quality. All provable good measurement, with random matrics in CS algorithm, can enhance speech signal. Objective evaluation on various dB SNR shows that the proposed algorithm exhibits better noise reduction ability over conventional approaches without obvious degradation of the speech signal quality.
|
|---|