ISCA Archive Interspeech 2009
ISCA Archive Interspeech 2009

Model based feature enhancement for automatic speech recognition in reverberant environments

Alexander Krueger, Reinhold Haeb-Umbach

In this paper we present a new feature space dereverberation technique for automatic speech recognition. We derive an expression for the dependence of the reverberant speech features in the log-mel spectral domain on the non-reverberant speech features and the room impulse response. The obtained observation model is used for a model based speech enhancement based on Kalman filtering. The performance of the proposed enhancement technique is studied on the AURORA5 database. In our currently best configuration, which includes uncertainty decoding, the number of recognition errors is approximately halved compared to the recognition of unprocessed speech.


doi: 10.21437/Interspeech.2009-355

Cite as: Krueger, A., Haeb-Umbach, R. (2009) Model based feature enhancement for automatic speech recognition in reverberant environments. Proc. Interspeech 2009, 1231-1234, doi: 10.21437/Interspeech.2009-355

@inproceedings{krueger09_interspeech,
  author={Alexander Krueger and Reinhold Haeb-Umbach},
  title={{Model based feature enhancement for automatic speech recognition in reverberant environments}},
  year=2009,
  booktitle={Proc. Interspeech 2009},
  pages={1231--1234},
  doi={10.21437/Interspeech.2009-355}
}