基于巴克谱的语音质量评估算法研究

提出了一种语音主观质量的客观评估算法，该算法在巴克谱的基础上计算原始语音与重建语音之间的失真度，并考虑了弱音帧与噪声帧的存在对语音质量评估的影响。文中同时给出了结合巴克谱失真和弱音与噪声帧比率的语音质量评估公式，并将计算结果与平均意见分(MOS)进行了比较。数值实验表明，本文提出的增强型巴克谱失真测度(IBSD)与MOS之间具有很强的相关性．能客观地评价出语音信号的主观质量，适用于各种语音编码、语音通信系统。

Algorithm for Bark Spectrum Based on Speech Quality Evaluation

An algorithm for objectively evaluating the subjective quality of speech is presented. The algorithm calculates the distortion between the original and the reconstructed signals in Bark spectrum domain and takes the effects into account which the muted and noise frames impose on the speech quality evaluation. This paper presents the formula of the speech quality evaluation incorporating the Bark spectrum distortion and the ratio of the muted and noise frames. Calculational results are comparied with the mean opinion score (MOS). Numerical experiments show that the improved Bark spectrum distortion (IBSD) measurement is highly correlated with MOS. IBSD can estimate the subjective quality of the speech signal. The IBSD can be used in various systems of speech coder and speech communication.