Speech enhancement for robust automatic speech recognition: Evaluation using a baseline system and instrumental measures
File(s)
Published version
Author(s)
Moore, AH
Peso, P
Naylor, PA
Type
Journal Article
Abstract
Automatic speech recognition in everyday environments must be robust to significant levels of reverberation and
noise. One strategy to achieve such robustness is multi-microphone speech enhancement. In this study, we present
results of an evaluation of different speech enhancement pipelines using a state-of-the-art
ASR
system for a wide
range of reverberation and noise conditions. The evaluation exploits the recently released ACE Challenge database
which includes measured multichannel acoustic impulse responses from 7 different rooms with reverberation times
ranging from 0.33 s to 1.34 s. The reverberant speech is mixed with ambient, fan and babble noise recordings made
with the same microphone setups in each of the rooms. In the first experiment performance of the
ASR
without
speech processing is evaluated. Results clearly indicate the deleterious effect of both noise and reverberation. In the
second experiment, different speech enhancement pipelines are evaluated with relative word error rate reductions of
up to 82%. Finally, the ability of selected instrumental metrics to predict
ASR
performance improvement is assessed.
The best performing metric, Short-Time Objective Intelligibility Measure, is shown to have a Pearson correlation
coefficient of 0.79, suggesting that it is a useful predictor of algorithm performance in these tests.
noise. One strategy to achieve such robustness is multi-microphone speech enhancement. In this study, we present
results of an evaluation of different speech enhancement pipelines using a state-of-the-art
ASR
system for a wide
range of reverberation and noise conditions. The evaluation exploits the recently released ACE Challenge database
which includes measured multichannel acoustic impulse responses from 7 different rooms with reverberation times
ranging from 0.33 s to 1.34 s. The reverberant speech is mixed with ambient, fan and babble noise recordings made
with the same microphone setups in each of the rooms. In the first experiment performance of the
ASR
without
speech processing is evaluated. Results clearly indicate the deleterious effect of both noise and reverberation. In the
second experiment, different speech enhancement pipelines are evaluated with relative word error rate reductions of
up to 82%. Finally, the ability of selected instrumental metrics to predict
ASR
performance improvement is assessed.
The best performing metric, Short-Time Objective Intelligibility Measure, is shown to have a Pearson correlation
coefficient of 0.79, suggesting that it is a useful predictor of algorithm performance in these tests.
Date Issued
2016-12-08
Date Acceptance
2016-11-25
Citation
Computer Speech and Language, 2016, 46, pp.574-584
ISSN
1095-8363
Publisher
Elsevier
Start Page
574
End Page
584
Journal / Book Title
Computer Speech and Language
Volume
46
Copyright Statement
© 2016 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license. (http://creativecommons.org/
licenses/by/4.0/)
licenses/by/4.0/)
Sponsor
Commission of the European Communities
Commission of the European Communities
Engineering & Physical Science Research Council (EPSRC)
Grant Number
PITN-GA-2012-316969
609465
ep/m026698/1
Subjects
Speech-Language Pathology & Audiology
0801 Artificial Intelligence And Image Processing
1702 Cognitive Science
Publication Status
Published