A Data-Driven Non-intrusive Measure of Speech Quality and Intelligibility
File(s)
Accepted version
Author(s)
Sharma, D
Naylor, PA
Wang, Y
Brookes, DM
Type
Journal Article
Abstract
Speech signals are often affected by additive noise
and distortion which can degrade the perceived quality and
intelligibility of the signal. We present a new measure, NISA, for
estimating the quality and intelligibility of speech degraded by
additive noise and distortions associated with telecommunications
networks, based on a data driven framework of feature extraction
and tree based regression. The new measure is non-intrusive,
operating on the degraded signal alone without the need for a
reference signal. This makes the measure applicable to practical
speech processing applications operating in the single-ended
mode. The new measure has been evaluated against the intrusive
measures PESQ and STOI. The results indicate that the accuracy
of the new non-intrusive method is around 90% of the accuracy of
the intrusive measures, depending on the test scenario. The NISA
measure therefore provides non-intrusive (single-ended) PESQ
and STOI estimates with high accuracy.
and distortion which can degrade the perceived quality and
intelligibility of the signal. We present a new measure, NISA, for
estimating the quality and intelligibility of speech degraded by
additive noise and distortions associated with telecommunications
networks, based on a data driven framework of feature extraction
and tree based regression. The new measure is non-intrusive,
operating on the degraded signal alone without the need for a
reference signal. This makes the measure applicable to practical
speech processing applications operating in the single-ended
mode. The new measure has been evaluated against the intrusive
measures PESQ and STOI. The results indicate that the accuracy
of the new non-intrusive method is around 90% of the accuracy of
the intrusive measures, depending on the test scenario. The NISA
measure therefore provides non-intrusive (single-ended) PESQ
and STOI estimates with high accuracy.
Date Issued
2016-04-26
Date Acceptance
2016-03-30
Citation
Speech Communication, 2016, 80, pp.84-94
ISSN
0167-6393
Publisher
Elsevier
Start Page
84
End Page
94
Journal / Book Title
Speech Communication
Volume
80
Copyright Statement
© 2016, Elsevier. Licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International http://creativecommons.org/licenses/by-nc-nd/4.0/
Sponsor
Home Office
Grant Number
PO 7073101
Subjects
Science & Technology
Technology
Acoustics
Computer Science, Interdisciplinary Applications
Computer Science
Speech quality
Speech intelligibility
CART
PESQ
STOI
PATTERN-RECOGNITION
NOISE
CLASSIFICATION
REVERBERANT
PERCEPTION
MODELS
Speech-Language Pathology & Audiology
0801 Artificial Intelligence And Image Processing
1702 Cognitive Science
2004 Linguistics
Publication Status
Published