基于听觉模型的子波变换语音增强

基于听觉模型的子波变换语音增强林宝成，富煜清黄志同（东南大学无线电工程系，南京２１００１８）（南京理工大学自控系，南京２１００１４）在许多实际的语音信号处理中，都迫切需要进行语音增强，例如，噪声环境中的语音识别［２］、语音编码、语音合成等。尽管有各种...

Speech Enhancement of Wavelet Transform Based on Auditory Model

This paper presents the speech enhancement principle of wavelet transform(WT)based on auditory medel.According to the mathematic representation of auditory physiological describtion,the ar-ticle gives a method of how to select WT and wavelet function and discusses the auditory denoising prin-ciple of this WT preeessing with mathematic analysis,So that the information processing machanism ofauditory system could be revealed deeply.Finally,serveral exporiments aredevised with the speech sig-nal corrupted by white noise,The results indicate the method has better ability for denoising.Themethod can be used in wide application because it can get both noise elimination and reduced signalspectrum simultaneously.