DocumentCode :
990604
Title :
On the use of channel-attentive MFCC for robust recognition of partially corrupted speech
Author :
Cho, Hoon-Young ; Oh, Yung-Hwan
Author_Institution :
Div. of Comput. Sci., Korea Adv. Inst. of Sci. & Technol., Daejeon, South Korea
Volume :
11
Issue :
6
fYear :
2004
fDate :
6/1/2004 12:00:00 AM
Firstpage :
581
Lastpage :
584
Abstract :
This letter proposes a channel-attentive mel frequency cepstral coefficient (CAMFCC) method to improve the utilization of uncorrupted or more reliable frequency bands for robust speech recognition. This method obtains a channel attention matrix by reliability estimation of mel filter bank channels, and both the input mel frequency cepstral coefficients and the mean vectors of hidden Markov models are corrected using the channel attention matrix at the output probability calculation of the Viterbi decoding. Experimental results on the TIDIGITS database corrupted by various band-selective noises indicated that the proposed CAMFCC method utilizes the uncorrupted partial frequency bands better than a multiband method, resolving the limitation of noise localization caused by the fixed boundaries of the multiband approach.
Keywords :
Viterbi decoding; cepstral analysis; channel bank filters; channel estimation; hidden Markov models; matrix algebra; speech recognition; Viterbi decoding; band-selective noise; channel attention matrix; channel-attentive Mel frequency cepstral coefficient method; hidden Markov models; mel filter bank channels estimation; partially corrupted speech recognition; Databases; Decoding; Filter bank; Frequency estimation; Hidden Markov models; Mel frequency cepstral coefficient; Probability; Robustness; Speech recognition; Viterbi algorithm; Band-selective noise; channel-attentive MFCC; multiband speech recognition; partially corrupted speech;
fLanguage :
English
Journal_Title :
Signal Processing Letters, IEEE
Publisher :
ieee
ISSN :
1070-9908
Type :
jour
DOI :
10.1109/LSP.2004.827922
Filename :
1300614
Link To Document :
بازگشت