Kallasjoki, H, Gemmeke, JF, Palomäki, KJ et al. (2 more authors) (2014) Recognition of Reverberant Speech by Missing Data Imputation and NMF Feature Enhancement. In: Proceedings of the REVERB Workshop 2014. REVERB Workshop, 10 May 2014, Florence, Italy. REVERB Challenge
Abstract
The problem of reverberation in speech recognition is addressed in this study by extending a noise-robust feature enhancement method based on non-negative matrix factorization. The signal model of the observation as a linear combination of sample spectrograms is augmented by a mel-spectral feature domain convolution to account for the effects of room reverberation. The proposed method is contrasted with missing data techniques for reverberant speech, and evaluated for speech recognition performance using the REVERB challenge corpus. Our results indicate consistent gains in recognition performance compared to the baseline system, with a relative improvement in word error rate of 42.6% for the optimal case.
Metadata
Item Type: | Proceedings Paper |
---|---|
Authors/Creators: |
|
Dates: |
|
Institution: | The University of Leeds |
Academic Units: | The University of Leeds > Faculty of Arts, Humanities and Cultures (Leeds) > School of Music (Leeds) |
Depositing User: | Symplectic Publications |
Date Deposited: | 26 Jun 2018 14:32 |
Last Modified: | 26 Jun 2018 14:32 |
Published Version: | https://reverb2014.dereverberation.com/proceedings... |
Status: | Published |
Publisher: | REVERB Challenge |
Open Archives Initiative ID (OAI ID): | oai:eprints.whiterose.ac.uk:118167 |