Deep-Learning-Based Audio-Visual Speech Enhancement in Presence of Lombard Effect

Research output: Contribution to journalJournal articleResearchpeer-review

Abstract

When speaking in presence of background noise, humans reflexively change their way of speaking in order to improve the intelligibility of their speech. This reflex is known as Lombard effect. Collecting speech in Lombard conditions is usually hard and costly. For this reason, speech enhancement systems are generally trained and evaluated on speech recorded in quiet to which noise is artificially added. Since these systems are often used in situations where Lombard speech occurs, in this work we perform an analysis of the impact that Lombard effect has on audio, visual and audio-visual speech enhancement, focusing on deep-learning-based systems, since they represent the current state of the art in the field.

We conduct several experiments using an audio-visual Lombard speech corpus consisting of utterances spoken by 54 different talkers. The results show that training deep-learning-based models with Lombard speech is beneficial in terms of both estimated speech quality and estimated speech intelligibility at low signal to noise ratios, where the visual modality can play an important role in acoustically challenging situations. We also find that a performance difference between genders exists due to the distinct Lombard speech exhibited by males and females, and we analyse it in relation with acoustic and visual features. Furthermore, listening tests conducted with audio-visual stimuli show that the speech quality of the signals processed with systems trained using Lombard speech is statistically significantly better than the one obtained using systems trained with non-Lombard speech at a signal to noise ratio of -5 dB. Regarding speech intelligibility, we find a general tendency of the benefit in training the systems with Lombard speech.
Original languageEnglish
JournalSpeech Communication
Volume115
Pages (from-to)38-50
Number of pages13
ISSN0167-6393
DOIs
Publication statusPublished - Dec 2019

Fingerprint

Speech Enhancement
Speech enhancement
learning
Speech intelligibility
Speech Intelligibility
Learning
Speech
Vision
Deep learning
Enhancement
Audiovisual Speech
Lombards
Signal to noise ratio
speaking
Modality
Acoustics

Keywords

  • Audio-visual speech enhancement
  • Deep learning
  • Lombard effect
  • Speech intelligibility
  • Speech quality

Cite this

@article{9b4c80fe33604b35b9de4690bfb0dcf1,
title = "Deep-Learning-Based Audio-Visual Speech Enhancement in Presence of Lombard Effect",
abstract = "When speaking in presence of background noise, humans reflexively change their way of speaking in order to improve the intelligibility of their speech. This reflex is known as Lombard effect. Collecting speech in Lombard conditions is usually hard and costly. For this reason, speech enhancement systems are generally trained and evaluated on speech recorded in quiet to which noise is artificially added. Since these systems are often used in situations where Lombard speech occurs, in this work we perform an analysis of the impact that Lombard effect has on audio, visual and audio-visual speech enhancement, focusing on deep-learning-based systems, since they represent the current state of the art in the field.We conduct several experiments using an audio-visual Lombard speech corpus consisting of utterances spoken by 54 different talkers. The results show that training deep-learning-based models with Lombard speech is beneficial in terms of both estimated speech quality and estimated speech intelligibility at low signal to noise ratios, where the visual modality can play an important role in acoustically challenging situations. We also find that a performance difference between genders exists due to the distinct Lombard speech exhibited by males and females, and we analyse it in relation with acoustic and visual features. Furthermore, listening tests conducted with audio-visual stimuli show that the speech quality of the signals processed with systems trained using Lombard speech is statistically significantly better than the one obtained using systems trained with non-Lombard speech at a signal to noise ratio of -5 dB. Regarding speech intelligibility, we find a general tendency of the benefit in training the systems with Lombard speech.",
keywords = "Audio-visual speech enhancement, Deep learning, Lombard effect, Speech intelligibility, Speech quality",
author = "Daniel Michelsanti and Zheng-Hua Tan and Sigurdur Sigurdsson and Jesper Jensen",
year = "2019",
month = "12",
doi = "10.1016/j.specom.2019.10.006",
language = "English",
volume = "115",
pages = "38--50",
journal = "Speech Communication",
issn = "0167-6393",
publisher = "Elsevier",

}

Deep-Learning-Based Audio-Visual Speech Enhancement in Presence of Lombard Effect. / Michelsanti, Daniel; Tan, Zheng-Hua; Sigurdsson, Sigurdur; Jensen, Jesper.

In: Speech Communication, Vol. 115, 12.2019, p. 38-50.

Research output: Contribution to journalJournal articleResearchpeer-review

TY - JOUR

T1 - Deep-Learning-Based Audio-Visual Speech Enhancement in Presence of Lombard Effect

AU - Michelsanti, Daniel

AU - Tan, Zheng-Hua

AU - Sigurdsson, Sigurdur

AU - Jensen, Jesper

PY - 2019/12

Y1 - 2019/12

N2 - When speaking in presence of background noise, humans reflexively change their way of speaking in order to improve the intelligibility of their speech. This reflex is known as Lombard effect. Collecting speech in Lombard conditions is usually hard and costly. For this reason, speech enhancement systems are generally trained and evaluated on speech recorded in quiet to which noise is artificially added. Since these systems are often used in situations where Lombard speech occurs, in this work we perform an analysis of the impact that Lombard effect has on audio, visual and audio-visual speech enhancement, focusing on deep-learning-based systems, since they represent the current state of the art in the field.We conduct several experiments using an audio-visual Lombard speech corpus consisting of utterances spoken by 54 different talkers. The results show that training deep-learning-based models with Lombard speech is beneficial in terms of both estimated speech quality and estimated speech intelligibility at low signal to noise ratios, where the visual modality can play an important role in acoustically challenging situations. We also find that a performance difference between genders exists due to the distinct Lombard speech exhibited by males and females, and we analyse it in relation with acoustic and visual features. Furthermore, listening tests conducted with audio-visual stimuli show that the speech quality of the signals processed with systems trained using Lombard speech is statistically significantly better than the one obtained using systems trained with non-Lombard speech at a signal to noise ratio of -5 dB. Regarding speech intelligibility, we find a general tendency of the benefit in training the systems with Lombard speech.

AB - When speaking in presence of background noise, humans reflexively change their way of speaking in order to improve the intelligibility of their speech. This reflex is known as Lombard effect. Collecting speech in Lombard conditions is usually hard and costly. For this reason, speech enhancement systems are generally trained and evaluated on speech recorded in quiet to which noise is artificially added. Since these systems are often used in situations where Lombard speech occurs, in this work we perform an analysis of the impact that Lombard effect has on audio, visual and audio-visual speech enhancement, focusing on deep-learning-based systems, since they represent the current state of the art in the field.We conduct several experiments using an audio-visual Lombard speech corpus consisting of utterances spoken by 54 different talkers. The results show that training deep-learning-based models with Lombard speech is beneficial in terms of both estimated speech quality and estimated speech intelligibility at low signal to noise ratios, where the visual modality can play an important role in acoustically challenging situations. We also find that a performance difference between genders exists due to the distinct Lombard speech exhibited by males and females, and we analyse it in relation with acoustic and visual features. Furthermore, listening tests conducted with audio-visual stimuli show that the speech quality of the signals processed with systems trained using Lombard speech is statistically significantly better than the one obtained using systems trained with non-Lombard speech at a signal to noise ratio of -5 dB. Regarding speech intelligibility, we find a general tendency of the benefit in training the systems with Lombard speech.

KW - Audio-visual speech enhancement

KW - Deep learning

KW - Lombard effect

KW - Speech intelligibility

KW - Speech quality

UR - http://www.scopus.com/inward/record.url?scp=85074407484&partnerID=8YFLogxK

U2 - 10.1016/j.specom.2019.10.006

DO - 10.1016/j.specom.2019.10.006

M3 - Journal article

VL - 115

SP - 38

EP - 50

JO - Speech Communication

JF - Speech Communication

SN - 0167-6393

ER -