

浏览全部资源
扫码关注微信
清华大学电子工程系微波与数字通信国家重点实验室
Published:1997
移动端阅览
[1]刘志勇,唐昆,冯重熙.一种高质量的8kb/s ACELP语音编码算法及其实时实现[J].电子学报,1997(07):72-74.
刘志勇, 唐昆, 冯重熙. A High Quality 8kb/s ACELP Coder and lts Real-Time Realization[J]. Acta Electronica Sinica, 1997, (7).
本文介绍了一种编码速率为8kb/s的高质量实时语音编码器.它采用了代数码本激励线性预测(ACELP)的编码方法,并采用高效的码本结构、码本搜索技术和矢量量化技术来获得较高的语音合成质量和较低的算法复杂度.在无需外部RAM和ROM的情况下,该算法已用TMS320C50实时实现并用于一个实时的全双工通信系统.通过信噪比及人耳主观听觉实验等性能测试表明,该算法的性能明显优于北美的8kb/sVSELP语音编码算法而算法的复杂度却与VSELP相当.
In this paper
a kind of real-time speech coder with high quality and low complexity at the bit rate of 8kb/s is presented. The encoding algorithm is based on algebraic code excited linear prediction (ACELP). Efficient codebook structure and codebook search strategies
along with efficient vector quantization procedures
have been developed to achieve high quality reconstructed speech while maintaining low complexity. Without any external RAM or ROM
the algorithm has been implemented on TMS320C50 for a realtime full duplex communication system.Through performance tests of SNR and subjective listening
we have verified that the proposed algorithm is better than the 8kb/s VSELP algorithm and its complexity is almost the same as that of VSELP algorithm.
0
Views
66
下载量
5
CSCD
Publicity Resources
Related Articles
Related Author
Related Institution
京公网安备11010802024621