| HAL-Inria | Publications, software ... of Inria's scientists |
Conference papers
An extended experimental investigation of DNN uncertainty propagation for noise robust ASR
Abstract : Automatic speech recognition (ASR) in noisy environments remains a challenging goal. Recently, the idea of estimating the uncertainty about the features obtained after speech enhancement and propagating it to dynamically adapt deep neural network (DNN) based acoustic models has raised some interest. However, the results in the literature were reported on simulated noisy datasets for a limited variety of uncertainty estimators. We found that they vary significantly in different conditions. Hence, the main contribution of this work is to assess DNN uncertainty decoding performance for different data conditions and different uncertainty estimation/propagation techniques. In addition, we propose a neural network based uncertainty estima-tor and compare it with other uncertainty estimators. We report detailed ASR results on the CHiME-2 and CHiME-3 datasets. We find that, on average, uncertainty propagation provides similar relative improvement on real and simulated data and that the proposed uncertainty estimator performs significantly better than the one in [1]. We also find that the improvement is consistent, but it depends on the signal-to-noise ratio (SNR) and the noise environment.
Cited literature [29 references]
https://hal.inria.fr/hal-01446441
Contributor : Emmanuel Vincent <>
Submitted on : Wednesday, January 25, 2017 - 11:59:18 PM
Last modification on : Monday, May 4, 2020 - 11:37:57 AM
Document(s) archivé(s) le : Wednesday, April 26, 2017 - 6:53:17 PM
Contributor : Emmanuel Vincent <>
Submitted on : Wednesday, January 25, 2017 - 11:59:18 PM
Last modification on : Monday, May 4, 2020 - 11:37:57 AM
Document(s) archivé(s) le : Wednesday, April 26, 2017 - 6:53:17 PM
File
nathwani_HSCMA17.pdf
Files produced by the author(s)