Discriminative uncertainty estimation for noise robust ASR - Archive ouverte HAL Access content directly
Conference Papers Year :

Discriminative uncertainty estimation for noise robust ASR

(1) , (2) , (2)
1
2

Abstract

We consider the problem of uncertainty estimation for noise-robust ASR. Existing uncertainty estimation techniques improve ASR accuracy but they still exhibit a gap compared to the use of oracle uncertainty. This comes partly from the highly non-linear feature transformation and from ad- ditional assumptions such as Gaussian distribution and independence between frequency bins in the spectral domain. In this paper, we propose a method to rescale the estimated feature-domain full uncertainty covariance matrix in a state-dependent fashion according to a discriminative criterion. The state-dependent and feature index-dependent scaling factors are learned from development data. Experimental evaluation on Track 1 of the 2nd CHiME challenge data shows that discriminative rescaling leads to better results than generative rescaling. Moreover, discriminative rescaling of the Wiener uncertainty estimator leads to 12% relative word error rate reduction compared to discriminative rescaling of the alternative estimator in [1]
Fichier principal
Vignette du fichier
Icassp2015_bmmi.pdf (217.26 Ko) Télécharger le fichier
Origin : Files produced by the author(s)
Loading...

Dates and versions

hal-01103969 , version 1 (16-01-2015)

Identifiers

  • HAL Id : hal-01103969 , version 1

Cite

Dung Tien Tran, Emmanuel Vincent, Denis Jouvet. Discriminative uncertainty estimation for noise robust ASR. 40th IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) 2015, Apr 2015, Brisbane, Queensland, Australia. ⟨hal-01103969⟩
172 View
279 Download

Share

Gmail Facebook Twitter LinkedIn More