Uncertainty decoding using a sampling strategy based on the eigenvalue decomposition

Conference contribution


Publication Details

Author(s): Hümmer C, Stadter P, Kellermann W
Editor(s): VDE
Publication year: 2016
Pages range: 362-366
ISBN: 978-3-8007-4275-2
Language: English


Abstract

Uncertainty decoding combines a probabilistic distortion model with the
acoustic model of a speech recognition system. This can be realized for
DNN-based acoustic models by drawing feature samples from an estimated
probability distribution and averaging the resulting set of posterior
likelihoods at the output of the DNN. According to this principle, we
consider a probabilistic feature description in the logmelspec domain to
model the front-end estimation errors produced by a coherence-based
Wiener filter. As main innovation with respect to previous work, we
employ a sampling strategy based on the eigenvalue decomposition to
capture (instead of neglect) the cross-correlations between the acoustic
features as part of the uncertainty decoding scheme. The experimental
results for real recordings provided by the REVERB Challenge task
highlight the effectiveness of this sampling strategy in improving the
recognition accuracy of a DNN-HMM hybrid system.


FAU Authors / FAU Editors

Hümmer, Christian
Professur für Nachrichtentechnik
Kellermann, Walter Prof. Dr.-Ing.
Professur für Nachrichtentechnik


How to cite

APA:
Hümmer, C., Stadter, P., & Kellermann, W. (2016). Uncertainty decoding using a sampling strategy based on the eigenvalue decomposition. In VDE (Eds.), Proceedings of the 12th ITG Symposium on Speech Communication (pp. 362-366). Paderborn, DE.

MLA:
Hümmer, Christian, Philipp Stadter, and Walter Kellermann. "Uncertainty decoding using a sampling strategy based on the eigenvalue decomposition." Proceedings of the 12th ITG Symposium on Speech Communication, Paderborn Ed. VDE, 2016. 362-366.

BibTeX: 

Last updated on 2019-19-04 at 13:38