On using the auditory image model and invariant-integration for noise robust automatic speech recognition

F. Müller; A. Mertins

doi:10.1109/ICASSP.2012.6289019

On using the auditory image model and invariant-integration for noise robust automatic speech recognition

Institute of Signal Processing

Abstract

Commonly used feature extraction methods for automatic speech recognition (ASR) incorporate only rudimentary psychoacoustic findings. Several works showed that a physiologically closer auditory processing during the feature extraction stage can enhance the robustness of an ASR system in noisy environments. The “auditory image model” (AIM) is such a more sophisticated computational model. In this work we show how invariant integration can be applied in the feature space given by the AIM, and we analyze the performance of the resulting features under noisy conditions on the Aurora-2 task. Furthermore, we show that previously presented features based on power-normalization and invariant integration benefit from the AIM-based integration features when the feature vectors are combined with each other.

Original language	English
Title of host publication	2012 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)
Number of pages	4
Publisher	IEEE
Publication date	01.03.2012
Pages	4905-4908
ISBN (Print)	978-1-4673-0045-2
ISBN (Electronic)	978-1-4673-0046-9
DOIs	https://doi.org/10.1109/ICASSP.2012.6289019
Publication status	Published - 01.03.2012
Event	2012 IEEE International Conference on Acoustics, Speech, and Signal Processing - Kyoto, Japan Duration: 25.03.2012 → 30.03.2012 Conference number: 93091

UN SDGs

This output contributes to the following UN Sustainable Development Goals (SDGs)

Access to Document

10.1109/ICASSP.2012.6289019

Cite this

Müller, F., & Mertins, A. (2012). On using the auditory image model and invariant-integration for noise robust automatic speech recognition. In 2012 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) (pp. 4905-4908). (IEEE International Conference on Acoustics, Speech and Signal Processing). IEEE. https://doi.org/10.1109/ICASSP.2012.6289019

@inproceedings{b9fca8054f844e78840930fb64c1c842,

title = "On using the auditory image model and invariant-integration for noise robust automatic speech recognition",

abstract = "Commonly used feature extraction methods for automatic speech recognition (ASR) incorporate only rudimentary psychoacoustic findings. Several works showed that a physiologically closer auditory processing during the feature extraction stage can enhance the robustness of an ASR system in noisy environments. The “auditory image model” (AIM) is such a more sophisticated computational model. In this work we show how invariant integration can be applied in the feature space given by the AIM, and we analyze the performance of the resulting features under noisy conditions on the Aurora-2 task. Furthermore, we show that previously presented features based on power-normalization and invariant integration benefit from the AIM-based integration features when the feature vectors are combined with each other.",

author = "F. M{\"u}ller and A. Mertins",

year = "2012",

month = mar,

day = "1",

doi = "10.1109/ICASSP.2012.6289019",

language = "English",

isbn = "978-1-4673-0045-2",

series = "IEEE International Conference on Acoustics, Speech and Signal Processing",

publisher = "IEEE",

pages = "4905--4908",

booktitle = "2012 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)",

note = "2012 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2012 ; Conference date: 25-03-2012 Through 30-03-2012",

}

Müller, F & Mertins, A 2012, On using the auditory image model and invariant-integration for noise robust automatic speech recognition. in 2012 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP). IEEE International Conference on Acoustics, Speech and Signal Processing, IEEE, pp. 4905-4908, 2012 IEEE International Conference on Acoustics, Speech, and Signal Processing, Kyoto, Japan, 25.03.12. https://doi.org/10.1109/ICASSP.2012.6289019

On using the auditory image model and invariant-integration for noise robust automatic speech recognition. / Müller, F.; Mertins, A.
2012 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP). IEEE, 2012. p. 4905-4908 (IEEE International Conference on Acoustics, Speech and Signal Processing).

Research output: Chapters in Books/Reports/Conference Proceedings › Conference contribution › peer-review

TY - GEN

T1 - On using the auditory image model and invariant-integration for noise robust automatic speech recognition

AU - Müller, F.

AU - Mertins, A.

N1 - Conference code: 93091

PY - 2012/3/1

Y1 - 2012/3/1

N2 - Commonly used feature extraction methods for automatic speech recognition (ASR) incorporate only rudimentary psychoacoustic findings. Several works showed that a physiologically closer auditory processing during the feature extraction stage can enhance the robustness of an ASR system in noisy environments. The “auditory image model” (AIM) is such a more sophisticated computational model. In this work we show how invariant integration can be applied in the feature space given by the AIM, and we analyze the performance of the resulting features under noisy conditions on the Aurora-2 task. Furthermore, we show that previously presented features based on power-normalization and invariant integration benefit from the AIM-based integration features when the feature vectors are combined with each other.

AB - Commonly used feature extraction methods for automatic speech recognition (ASR) incorporate only rudimentary psychoacoustic findings. Several works showed that a physiologically closer auditory processing during the feature extraction stage can enhance the robustness of an ASR system in noisy environments. The “auditory image model” (AIM) is such a more sophisticated computational model. In this work we show how invariant integration can be applied in the feature space given by the AIM, and we analyze the performance of the resulting features under noisy conditions on the Aurora-2 task. Furthermore, we show that previously presented features based on power-normalization and invariant integration benefit from the AIM-based integration features when the feature vectors are combined with each other.

UR - https://www.scopus.com/record/display.uri?eid=2-s2.0-84867609306&origin=inward&txGid=d3ae7f9d62748c0fd519748607c514d0

U2 - 10.1109/ICASSP.2012.6289019

DO - 10.1109/ICASSP.2012.6289019

M3 - Conference contribution

SN - 978-1-4673-0045-2

T3 - IEEE International Conference on Acoustics, Speech and Signal Processing

SP - 4905

EP - 4908

BT - 2012 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)

PB - IEEE

T2 - 2012 IEEE International Conference on Acoustics, Speech, and Signal Processing

Y2 - 25 March 2012 through 30 March 2012

ER -

On using the auditory image model and invariant-integration for noise robust automatic speech recognition

Abstract

UN SDGs

Access to Document

Other files and links

Fingerprint

Invariant features for automatic speech recognition based on complex models of speech production and auditory perception

Cite this

On using the auditory image model and invariant-integration for noise robust automatic speech recognition

Abstract

UN SDGs

Access to Document

Other files and links

Fingerprint

Projects

Invariant features for automatic speech recognition based on complex models of speech production and auditory perception

Cite this