论文部分内容阅读
提出了一种基于听觉特性的语音失真测度方法──感知谱失真 PSD(Perceptual Spectrum Distortion)测度,该测度方法通过模拟人的听觉特性把语音短时频谱转变为符合听觉特性的感知频谱,再以感知谱为基础来度量语音失真程度。经过对不同质量的语音进行仿真实验以及与Itakura测度方法作对比实验,结果表明PSD测度是一种与语音质量主观评价一致性较好的语音失真测度方法。
This paper proposes a Perceptual Spectrum Distortion (PSD) measurement method based on auditory characteristics, which transforms the speech short-time spectrum into a perceptual spectrum conforming to auditory characteristics by simulating human auditory characteristics. Perceptual Spectrum is a measure of the level of speech distortion. After simulating the speech of different quality and comparing with the Itakura measure method, the result shows that the PSD measure is a measure method of speech distortion which is in good agreement with the subjective evaluation of speech quality.