听觉模型与语音信号处理方法的研究

在通信领域中,语音编码是语音信号处理的重要分支。为了适合信道传输,语音必须变换形式,基于承载信息并且保留信号,尽可能地处理。在当今的通信、计算机网络等应用领域中,具备低延迟、低码率两大特性的语音编码算法,发挥着决定性作用。在语音编码中,线性预测分析技术主要应用在感觉加权滤波器、综合滤波器及对数增益滤波器,该技术发挥着关键作用。文中的工作是呈现出一种混合LPC（Auditory-Acoustic-Hybrid-LPC）系数,它结合声学特性与听觉特性,以便提高编码后合成语音的听觉质量,这对编码算法的钻研有积极意义。

Research on Auditory Model and Speech Signal Processing Method

In the field of communication,speech coding is an important branch of speech signal processing.It reserves the signal carrying the information,based on the maximum degree of its processing,it transforms into a suitable form of transmission channels.In the today＇s society,speech coding algorithms with low bit rate and low delay,plays a key role in the communication,computer networks and many other applications.In the speech coding,feeling weighted filters,synthesis filters,logarithmic gain filters are related to the linear prediction analysis techniques,which plays a central role in the voice processing.The main work of this paper is to propose a combination of acoustic features and auditory features of the mixed LPC coefficients（Auditory-Acoustic-Hybrid-LPC）,making the audio quality of the encoded synthetic voice improved,which has a positive meaning to the research of coding algorithm.