Skip to Main content Skip to Navigation
Preprints, Working Papers, ...

Nonparametric uncertainty estimation and propagation for noise robust ASR

Dung T. Tran 1 Emmanuel Vincent 1 Denis Jouvet 1
1 PAROLE - Analysis, perception and recognition of speech
Inria Nancy - Grand Est, LORIA - NLPKD - Department of Natural Language Processing & Knowledge Discovery
Abstract : We consider the framework of uncertainty propa-gation for automatic speech recognition (ASR) in highly non-stationary noise environments. Uncertainty is considered as the variance of speech distortion. Yet, its accurate estimation in the spectral domain and its propagation to the feature domain remain difficult. Existing methods typically rely on a single uncertainty estimator and propagator fixed by mathematical approximation. In this paper, we propose a new paradigm where we seek to learn more powerful mappings to predict uncertainty from data. We investigate two such possible mappings: linear fusion of multiple uncertainty estimators/propagators and nonparametric uncertainty estimation/propagation. In addition, a procedure to propagate the estimated spectral-domain uncertainty to the static Mel frequency cepstral coefficients (MFCCs), to the log-energy, and to their first-and second-order time derivatives is proposed. This results in a full uncertainty covariance matrix over both static and dynamic MFCCs. Experimental evaluation on Track 1 of the 2nd CHiME Challenge corpus resulted in up to 29% relative keyword error rate reduction with respect to speech enhancement alone.
Document type :
Preprints, Working Papers, ...
Complete list of metadatas
Contributor : Emmanuel Vincent <>
Submitted on : Monday, February 9, 2015 - 11:32:40 AM
Last modification on : Tuesday, December 18, 2018 - 4:38:02 PM
Long-term archiving on: : Saturday, September 12, 2015 - 9:51:07 AM


Files produced by the author(s)


  • HAL Id : hal-01114329, version 1


Dung T. Tran, Emmanuel Vincent, Denis Jouvet. Nonparametric uncertainty estimation and propagation for noise robust ASR. 2015. ⟨hal-01114329v1⟩



Record views


Files downloads