ISCA Archive Interspeech 2013
ISCA Archive Interspeech 2013

Attribute-based histogram equalization (HEQ) and its adaptation for robust speech recognition

Xiong Xiao, Eng Siong Chng, Haizhou Li

Histogram equalization (HEQ) is a simple and effective feature normalization technique for robust speech recognition. Recently, we proposed to adapt HEQ transform to each test utterance using a maximum likelihood (ML) criterion and observed improved performance. In this paper, we further the study by applying attribute-based HEQ and its ML adaptation. Instead of applying a global HEQ transform to the test utterance, we propose to apply different HEQ transforms to the 6 manners of speech, e.g. vowel and fricative. We also developed the ML adaptation algorithm of the attribute-based HEQ. Experimental results show that the attribute-based HEQ adaptation obtained 21.8% and 19.5% relative error rate reduction over the global HEQ baseline on the Aurora-2 and Aurora-4 benchmarking tasks, respectively.


doi: 10.21437/Interspeech.2013-259

Cite as: Xiao, X., Chng, E.S., Li, H. (2013) Attribute-based histogram equalization (HEQ) and its adaptation for robust speech recognition. Proc. Interspeech 2013, 876-880, doi: 10.21437/Interspeech.2013-259

@inproceedings{xiao13b_interspeech,
  author={Xiong Xiao and Eng Siong Chng and Haizhou Li},
  title={{Attribute-based histogram equalization (HEQ) and its adaptation for robust speech recognition}},
  year=2013,
  booktitle={Proc. Interspeech 2013},
  pages={876--880},
  doi={10.21437/Interspeech.2013-259}
}