For Full-Text PDF, please login, if you are a member of IEICE,|
or go to Pay Per View on menu list, if you are a nonmember of IEICE.
Computationally Efficient Cepstral Domain Feature Compensation
Woohyung LIM Chang Woo HAN Nam Soo KIM
IEICE TRANSACTIONS on Information and Systems
Publication Date: 2009/01/01
Online ISSN: 1745-1361
Print ISSN: 0916-8532
Type of Manuscript: LETTER
Category: Speech and Hearing
feature compensation, cepstral domain, linear approximation,
Full Text: PDF(90.7KB)>>
In this letter, we propose a novel approach to feature compensation performed in the cepstral domain. Processing in the cepstral domain has the advantage that the spectral correlation among different frequencies is taken into consideration. By introducing a linear approximation with diagonal covariance assumption, we modify the conventional log-spectral domain feature compensation technique to fit to the cepstral domain. The proposed approach shows significant improvements in the AURORA2 speech recognition task.