An intelligibility metric based on a simple model of speech communication

Steven van Kuyk; W. Bastiaan Kleijn; Richard C. Hendriks

doi:10.1109/iwaenc.2016.7602933

An intelligibility metric based on a simple model of speech communication

Steven van Kuyk, W. Bastiaan Kleijn, Richard C. Hendriks

Signal Processing Systems

Research output: Chapter in Book/Conference proceedings/Edited volume › Conference contribution › Scientific › peer-review

5 Citations (Scopus)

78 Downloads (Pure)

Abstract

Instrumental measures of speech intelligibility typically produce an index between 0 and 1 that is monotonically related to listening test scores. As such, these measures are dimensionless and do not represent physical quantities. In this paper, we propose a new instrumental intelligibility metric that describes speech intelligibility using bits per second. The proposed metric builds upon an existing intelligibility metric that was motivated by information theory. Our main contribution is that we use a statistical model of speech communication that accounts for noise inherent in the speech production process. Experiments show that the proposed metric performs at least as well as existing state-of-the-art intelligibility metrics.

Original language	English
Title of host publication	2016 IEEE International Workshop on Acoustic Signal Enhancement (IWAENC)
Place of Publication	Piscataway, NJ
Publisher	IEEE
Pages	1-5
Number of pages	5
ISBN (Electronic)	978-1-5090-2007-2
DOIs	https://doi.org/10.1109/iwaenc.2016.7602933
Publication status	Published - 24 Oct 2016
Event	2016 IEEE International Workshop on Acoustic Signal Enhancement: IWAENC - Xi'an, China Duration: 13 Sept 2016 → 16 Sept 2016 http://www.iwaenc2016.org/

Conference

Conference	2016 IEEE International Workshop on Acoustic Signal Enhancement
Country/Territory	China
City	Xi'an
Period	13/09/16 → 16/09/16
Internet address	http://www.iwaenc2016.org/

Bibliographical note

(Best student paper award)

Keywords

mutual information
Intelligibility

Access to Document

10.1109/iwaenc.2016.7602933

kuyk_iwaenc16Accepted author manuscript, 237 KB

Cite this

@inproceedings{d3f5ef821a60432e86c81280014a7866,

title = "An intelligibility metric based on a simple model of speech communication",

abstract = "Instrumental measures of speech intelligibility typically produce an index between 0 and 1 that is monotonically related to listening test scores. As such, these measures are dimensionless and do not represent physical quantities. In this paper, we propose a new instrumental intelligibility metric that describes speech intelligibility using bits per second. The proposed metric builds upon an existing intelligibility metric that was motivated by information theory. Our main contribution is that we use a statistical model of speech communication that accounts for noise inherent in the speech production process. Experiments show that the proposed metric performs at least as well as existing state-of-the-art intelligibility metrics.",

keywords = "mutual information, Intelligibility",

author = "{van Kuyk}, Steven and Kleijn, {W. Bastiaan} and Hendriks, {Richard C.}",

note = "(Best student paper award); 2016 IEEE International Workshop on Acoustic Signal Enhancement : IWAENC ; Conference date: 13-09-2016 Through 16-09-2016",

year = "2016",

month = oct,

day = "24",

doi = "10.1109/iwaenc.2016.7602933",

language = "English",

pages = "1--5",

booktitle = "2016 IEEE International Workshop on Acoustic Signal Enhancement (IWAENC)",

publisher = "IEEE",

address = "United States",

url = "http://www.iwaenc2016.org/",

}

An intelligibility metric based on a simple model of speech communication. / van Kuyk, Steven; Kleijn, W. Bastiaan ; Hendriks, Richard C.
2016 IEEE International Workshop on Acoustic Signal Enhancement (IWAENC). Piscataway, NJ: IEEE, 2016. p. 1-5.

Research output: Chapter in Book/Conference proceedings/Edited volume › Conference contribution › Scientific › peer-review

TY - GEN

T1 - An intelligibility metric based on a simple model of speech communication

AU - van Kuyk, Steven

AU - Kleijn, W. Bastiaan

AU - Hendriks, Richard C.

N1 - (Best student paper award)

PY - 2016/10/24

Y1 - 2016/10/24

N2 - Instrumental measures of speech intelligibility typically produce an index between 0 and 1 that is monotonically related to listening test scores. As such, these measures are dimensionless and do not represent physical quantities. In this paper, we propose a new instrumental intelligibility metric that describes speech intelligibility using bits per second. The proposed metric builds upon an existing intelligibility metric that was motivated by information theory. Our main contribution is that we use a statistical model of speech communication that accounts for noise inherent in the speech production process. Experiments show that the proposed metric performs at least as well as existing state-of-the-art intelligibility metrics.

AB - Instrumental measures of speech intelligibility typically produce an index between 0 and 1 that is monotonically related to listening test scores. As such, these measures are dimensionless and do not represent physical quantities. In this paper, we propose a new instrumental intelligibility metric that describes speech intelligibility using bits per second. The proposed metric builds upon an existing intelligibility metric that was motivated by information theory. Our main contribution is that we use a statistical model of speech communication that accounts for noise inherent in the speech production process. Experiments show that the proposed metric performs at least as well as existing state-of-the-art intelligibility metrics.

KW - mutual information

KW - Intelligibility

UR - http://resolver.tudelft.nl/uuid:d3f5ef82-1a60-432e-86c8-1280014a7866

U2 - 10.1109/iwaenc.2016.7602933

DO - 10.1109/iwaenc.2016.7602933

M3 - Conference contribution

SP - 1

EP - 5

BT - 2016 IEEE International Workshop on Acoustic Signal Enhancement (IWAENC)

PB - IEEE

CY - Piscataway, NJ

T2 - 2016 IEEE International Workshop on Acoustic Signal Enhancement

Y2 - 13 September 2016 through 16 September 2016

ER -

An intelligibility metric based on a simple model of speech communication

Abstract

Conference

Bibliographical note

Keywords

Access to Document

Other files and links

Fingerprint

Cite this