Sixth International Conference on Spoken Language Processing
(ICSLP 2000)

Beijing, China
October 16-20, 2000

On the Importance of Components of the MFCC in Speech and Speaker Recognition

Bin Zhen, Xihong Wu, Zhimin Liu, Huisheng Chi

Center for Information Science, Peking University, Beijing, China

In this paper, we analyzed the relative importance of components of MFCC for both speech recognition and speaker recognition using DTW recognizer in various noise environments. For English digit and under the Euclidean distance definition, the experiment results show cepstral components from C2 to C16 contain the most useful speaker information, while C0 and C1 are usually harm to speaker recognition. Cepstral terms from C1 to C12 are found to contain the most useful speech information. In both tasks, the additive noise decreases the relative importance of low MFCC terms faster than that of the middle and high MFCC terms, and the decrement depends on the speech SNR. The channel distortion will deteriorate low terms more than the middle and high MFCC terms in both tasks, also.

Full Paper

Bibliographic reference.  Zhen, Bin / Wu, Xihong / Liu, Zhimin / Chi, Huisheng (2000): "On the importance of components of the MFCC in speech and speaker recognition", In ICSLP-2000, vol.2, 487-490.