首页 | 本学科首页   官方微博 | 高级检索  
     

一种改进型自适应前后向量化IMBE算法
引用本文:李烨,杨家玮,梁彦霞.一种改进型自适应前后向量化IMBE算法[J].西安电子科技大学学报,2009,36(6):968-971+984.
作者姓名:李烨  杨家玮  梁彦霞
作者单位:(西安电子科技大学 综合业务网理论及关键技术国家重点实验室,陕西 西安710071)
基金项目:国家杰出青年科学基金资助,长江学者和创新团队发展计划资助,高等学校创新引智计划资助 
摘    要:为了提高自适应后向码本的精度,提出了一种码本更新的新算法.该算法设置了一个线谱频率系数自适应后向码本,通过对该码本中最新一个码矢量与当前帧线谱频率系数内插来提高线性预测系数自适应后向码本的精度,从而提高了后向量化的使用率,降低了平均编码速率.主客观测试表明,改进算法的合成语音质量好于自适应前后向量化IMBE算法,且合成语音质量与 2.4kb/s IMBE算法相当.

关 键 词:语音编码  多带激励  MBE  量化  内插  
收稿时间:2009-04-15

Improved IMBE algorithm with adaptive forward-backward quantization
LI Ye,YANG Jia-wei,LIANG Yan-xia.Improved IMBE algorithm with adaptive forward-backward quantization[J].Journal of Xidian University,2009,36(6):968-971+984.
Authors:LI Ye  YANG Jia-wei  LIANG Yan-xia
Affiliation:(State Key Lab. of Integrated Service Networks, Xidian Univ., Xi'an  710071, China)
Abstract:A new method for updating the codebook is proposed to improve the accuracy of the adaptive backward codebook. This new algorithm sets up a line spectrum frequency coefficient's adaptive backward codebook, which implements the interpolation between the line spectrum frequency coefficients to improve the precision of the linear prediction coefficient's adaptive backward codebook. As a result, it increases the use frequency of the backward quantization and reduces the average coding rate. Objective and subjective tests show that the quality of speech synthesized by the improved method is superior to the one synthesized by the IMBE algorithm with adaptive forward-backward quantization, and as good as the one synthesized by the 2.4kbps IMBE algorithm.
Keywords:speech coding  multi-band excitation(MBE)  quantization  interpolation
本文献已被 万方数据 等数据库收录!
点击此处可从《西安电子科技大学学报》浏览原始摘要信息
点击此处可从《西安电子科技大学学报》下载全文
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号