SELP低速率语音编码算法优化及实现
首发时间:2012-12-31
摘要:本文首先简单介绍了清华大学提出的正弦激励线性预测(SELP)算法与目前主流的几种低速率语音编码算法相比具有的优势,着重分析了SELP算法的基本原理,结合混合激励线性预测(MELP)算法的关键技术及其多帧联合、参数内插等降低编码速率的方法,对SELP算法进行了改进,提出了1.2Kb/s SELP算法,并在PC机上实现了此算法,从而将两种速率的SELP算法性能进行比较,对于语音编码算法研究者具有一定的参考价值。
For information in English, please click here
Optimization and Implementation for Sinusoidal Excitation Linear Prediction Speech Coding Algorithm
Abstract:After briefly introduces the advantages of the SELP algorithm comparing with several currently excellent low-rate speech coding algorithms, the paper analyzes the basic principle of the SELP algorithm.Combining the methods to reduce the coding rate in MELP algorithm, designs a 1.2Kb / s SELP algorithm and implement the algorithm on PC, then make a comparison of the two different rate algorithms, which has a reference value for researchers of speech coding algorithms.
Keywords: speech coding SELP dynamic programming parameters interpolation
基金:
论文图表:
引用
No.****
同行评议
共计0人参与
勘误表
SELP低速率语音编码算法优化及实现
评论
全部评论0/1000