DocumentCode :
3431520
Title :
Constrained adaptation of histogram equalization for robust speech recognition
Author :
Xiong Xiao ; Eng Siong Chng ; Haizhou Li
Author_Institution :
Temasek Lab., Nanyang Technol. Univ., Singapore, Singapore
fYear :
2013
fDate :
6-10 July 2013
Firstpage :
360
Lastpage :
364
Abstract :
Parametric histogram equalization (PHEQ) is a simple and effective way to reduce noise and channel effects in automatic speech recognition. Recently, we proposed to adapt the PHEQ transform towards individual test utterance to further improve the performance of PHEQ. In this paper, we improve the adaptation of PHEQ by exploiting the correlation between speech feature elements. Previously, PHEQ is applied to each feature element independently. In the case of Mel-frequency cepstral features (MFCC), this may not be optimal as it ignores the fact that the delta and acceleration features are derived from the static features and they are correlated. In this paper, we proposed to maintain the relationship between static, delta, and acceleration features to impose a constraint in PHEQ adaptation such that the PHEQ parameters will be more reliably estimated for short utterances. Experiments on both Aurora-2 and Aurora-4 tasks confirmed that it is beneficial to maintain the relationship between feature elements especially for short sentences.
Keywords :
equalisers; signal denoising; speech recognition; MFCC; Mel-frequency cepstral features; PHEQ transform; acceleration features; automatic speech recognition; channel effects; delta features; noise reduction; parametric histogram equalization; robust speech recognition; speech feature elements; test utterance; Acceleration; Hidden Markov models; Maximum likelihood estimation; Mel frequency cepstral coefficient; Speech; Speech recognition; Vectors; feature normalization; histogram equalization; maximum likelihood; robust ASR;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Signal and Information Processing (ChinaSIP), 2013 IEEE China Summit & International Conference on
Conference_Location :
Beijing
Type :
conf
DOI :
10.1109/ChinaSIP.2013.6625361
Filename :
6625361
Link To Document :
بازگشت