The speech cepstral features are very important parameter in ASR, which symbolizes the property of HAS(human auditory system). The paper discusses cepstral features based on linear predictive and non-linear MEL scale, and presents LPCC and MFCC's extraction algorithm and differential coefficient and accelerated coefficient are also presented. The recognition experiment results show that MFCC's robustness is prior to LPCCs.
| 
推荐 
点击下列相关研究机构和相关文献作者，可以直接查到这些机构和作者被《中国知识资源总库》收录的其它文献，使您全面了解该机构和该作者的研究动态和历史。
