英文摘要 |
We proposed an efficient method of converting speech codec formats between the G.729 and G.723.1 speech codec. The G.729 three frames is converted to one frame of the G.723.1, transcoding is completed through four processes: LSP and pitch conversions are used to linear interpolation processing, respectively, fast adaptive-codebook search processing adopt to predict the range of adaptive-codebook gain in the G.723.1, and a fast stochastic excitation pulses estimates method. Simulation results show that the proposed method is able to reduce the computation complexity about 84.87% with shorter delay than the tandem. And the perceptual speech quality (PESQ) evaluated better than the latter. |