Speech Breathing Estimation Using Deep Learning Methods

Venkata Srikanth Nallanthighal; Aki Harma; Helmer Strik

doi:10.1109/ICASSP40776.2020.9053753

Speech Breathing Estimation Using Deep Learning Methods

Venkata Srikanth Nallanthighal^*, Aki Harma, Helmer Strik

^*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceeding › Conference article in proceeding › Academic › peer-review

Abstract

Breathing is the primary mechanism for maintaining the subglottal pressure for speech production. Speech can be seen as a systematic outflow of air during exhalation characterized by linguistic content and prosodic factors. Thus, sensing respiratory dynamics from the speech is plausible. In this paper, we explore techniques for sensing breathing from speech using deep learning architectures including multi-task learning approaches. Estimating the breathing pattern from the speech would give us information about the respiration rate, breathing capacity and thus enable us to understand the pathological condition of a person using one's speech. Training and evaluation of our model on our database of breathing signal and speech for 40 subjects yielded a sensitivity of 0.88 for breath event detection and 5.6 % error for breathing rate estimation.

Original language	English
Title of host publication	2020 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2020 - Proceedings
Publisher	IEEE
Pages	1140-1144
Number of pages	5
ISBN (Electronic)	9781509066315
DOIs	https://doi.org/10.1109/ICASSP40776.2020.9053753
Publication status	Published - May 2020
Externally published	Yes
Event	45th International Conference on Acoustics, Speech, and Signal Processing - Online, Barcelona, Spain Duration: 4 May 2020 → 8 May 2020 Conference number: 45 https://2020.ieeeicassp.org/

Publication series

Series	ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings
Volume	2020-May
ISSN	1520-6149

Conference

Conference	45th International Conference on Acoustics, Speech, and Signal Processing
Abbreviated title	ICASSP 2020
Country/Territory	Spain
City	Barcelona
Period	4/05/20 → 8/05/20
Internet address	https://2020.ieeeicassp.org/

Keywords

deep neural networks
Multi task learning
signal processing
Speech breathing
speech technology

Access to Document

10.1109/ICASSP40776.2020.9053753

Cite this

@inproceedings{b4f69fedcd804fed8b844bafcfb7912d,

title = "Speech Breathing Estimation Using Deep Learning Methods",

abstract = "Breathing is the primary mechanism for maintaining the subglottal pressure for speech production. Speech can be seen as a systematic outflow of air during exhalation characterized by linguistic content and prosodic factors. Thus, sensing respiratory dynamics from the speech is plausible. In this paper, we explore techniques for sensing breathing from speech using deep learning architectures including multi-task learning approaches. Estimating the breathing pattern from the speech would give us information about the respiration rate, breathing capacity and thus enable us to understand the pathological condition of a person using one's speech. Training and evaluation of our model on our database of breathing signal and speech for 40 subjects yielded a sensitivity of 0.88 for breath event detection and 5.6 % error for breathing rate estimation.",

keywords = "deep neural networks, Multi task learning, signal processing, Speech breathing, speech technology",

author = "Nallanthighal, {Venkata Srikanth} and Aki Harma and Helmer Strik",

note = "Funding Information: This work was partially supported by the Horizon H2020 Marie Skodowska-Curie Actions Initial Training Network European Training Network project under grant agreement No. 766287 (TAPAS) and Data Science Department, Philips Research, Eindhoven. Publisher Copyright: {\textcopyright} 2020 IEEE.; 45th International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2020 ; Conference date: 04-05-2020 Through 08-05-2020",

year = "2020",

month = may,

doi = "10.1109/ICASSP40776.2020.9053753",

language = "English",

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

publisher = "IEEE",

pages = "1140--1144",

booktitle = "2020 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2020 - Proceedings",

address = "United States",

url = "https://2020.ieeeicassp.org/",

}

Nallanthighal, VS, Harma, A & Strik, H 2020, Speech Breathing Estimation Using Deep Learning Methods. in 2020 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2020 - Proceedings., 9053753, IEEE, ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings, vol. 2020-May, pp. 1140-1144, 45th International Conference on Acoustics, Speech, and Signal Processing, Barcelona, Spain, 4/05/20. https://doi.org/10.1109/ICASSP40776.2020.9053753

Speech Breathing Estimation Using Deep Learning Methods. / Nallanthighal, Venkata Srikanth; Harma, Aki; Strik, Helmer.
2020 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2020 - Proceedings. IEEE, 2020. p. 1140-1144 9053753 (ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings, Vol. 2020-May).

Research output: Chapter in Book/Report/Conference proceeding › Conference article in proceeding › Academic › peer-review

TY - GEN

T1 - Speech Breathing Estimation Using Deep Learning Methods

AU - Nallanthighal, Venkata Srikanth

AU - Harma, Aki

AU - Strik, Helmer

N1 - Conference code: 45

PY - 2020/5

Y1 - 2020/5

N2 - Breathing is the primary mechanism for maintaining the subglottal pressure for speech production. Speech can be seen as a systematic outflow of air during exhalation characterized by linguistic content and prosodic factors. Thus, sensing respiratory dynamics from the speech is plausible. In this paper, we explore techniques for sensing breathing from speech using deep learning architectures including multi-task learning approaches. Estimating the breathing pattern from the speech would give us information about the respiration rate, breathing capacity and thus enable us to understand the pathological condition of a person using one's speech. Training and evaluation of our model on our database of breathing signal and speech for 40 subjects yielded a sensitivity of 0.88 for breath event detection and 5.6 % error for breathing rate estimation.

AB - Breathing is the primary mechanism for maintaining the subglottal pressure for speech production. Speech can be seen as a systematic outflow of air during exhalation characterized by linguistic content and prosodic factors. Thus, sensing respiratory dynamics from the speech is plausible. In this paper, we explore techniques for sensing breathing from speech using deep learning architectures including multi-task learning approaches. Estimating the breathing pattern from the speech would give us information about the respiration rate, breathing capacity and thus enable us to understand the pathological condition of a person using one's speech. Training and evaluation of our model on our database of breathing signal and speech for 40 subjects yielded a sensitivity of 0.88 for breath event detection and 5.6 % error for breathing rate estimation.

KW - deep neural networks

KW - Multi task learning

KW - signal processing

KW - Speech breathing

KW - speech technology

UR - http://www.scopus.com/inward/record.url?scp=85089214829&partnerID=8YFLogxK

U2 - 10.1109/ICASSP40776.2020.9053753

DO - 10.1109/ICASSP40776.2020.9053753

M3 - Conference article in proceeding

AN - SCOPUS:85089214829

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

SP - 1140

EP - 1144

BT - 2020 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2020 - Proceedings

PB - IEEE

T2 - 45th International Conference on Acoustics, Speech, and Signal Processing

Y2 - 4 May 2020 through 8 May 2020

ER -

Nallanthighal VS, Harma A, Strik H. Speech Breathing Estimation Using Deep Learning Methods. In 2020 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2020 - Proceedings. IEEE. 2020. p. 1140-1144. 9053753. (ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings, Vol. 2020-May). doi: 10.1109/ICASSP40776.2020.9053753

Speech Breathing Estimation Using Deep Learning Methods

Abstract

Publication series

Conference

Keywords

Access to Document

Other files and links

Fingerprint

Cite this