This paper presents a new approach to improve the robustness of large vocabulary continuous speech recognition. The proposed technique { based on Singular Value Decomposition (SVD) { originates from classical signal enhancement, but it is adapted to the speci c requirements imposed by the speech recognition process. Additive noise reduction is obtained by altering the singular value spectrum of the signal observation matrix, thereby preserving speech signal components and suppressing noise-related components. The basic algorithms are developed for white noise but they can easily be extended to the general coloured noise case. With the aid of a noise reference, non-stationary noise can be handled as well. All schemes are adaptive, and work in real-time. Recognition experiments on a noise-corrupted database with large vocabulary, continuous speech (Resource Management) reveal that relative reductions of the WER of more than 60% are obtained.
[1]
Søren Holdt Jensen,et al.
FIR filter representations of reduced-rank noise reduction
,
1998,
IEEE Trans. Signal Process..
[2]
Dirk Van Compernolle,et al.
Fast and accurate acoustic modelling with semi-continuous HMMs
,
1998,
Speech Commun..
[3]
Fei Xie,et al.
Spectral estimation and normalisation for robust speech recognition
,
1996,
Proceeding of Fourth International Conference on Spoken Language Processing. ICSLP '96.
[4]
Yifan Gong,et al.
Speech recognition in noisy environments: A survey
,
1995,
Speech Commun..
[5]
Søren Holdt Jensen,et al.
Reduction of broad-band noise in speech by truncated QSVD
,
1995,
IEEE Trans. Speech Audio Process..