摘要
介绍了8kHz采样率下算法延迟为2.5ms的10kb/s语音编码算法。本算法使用20维的CELP,并用自适应码本搜索和代数码本搜索相级联的方法进行矢量量化。通过PC仿真实验并经PESQ测试,其合成语音质量接近16kb/s的G.728。
This paper introduces a 10 k/s speech coding algorithm with delay of 2.5 ms in sampling rate of 8kHz, which uses a CELP with frame size of 20 and a method of cascading adaptive codebook search and algebra codebook search to make vector quantization. The PC simulation and PESQ test show that the synthesized speech quality is close to G.728 of 16k/s.
出处
《科技情报开发与经济》
2011年第31期137-139,共3页
Sci-Tech Information Development & Economy