Emotion Recognition from Raw Speech Signals Using 2D CNN with Deep Metric Learning

Bogdan Mocanu; Ruxandra Tapu

doi:10.1109/ICCE53296.2022.9730534

Emotion Recognition from Raw Speech Signals Using 2D CNN with Deep Metric Learning

Bogdan Mocanu
, Ruxandra Tapu

Telecom SudParis

University 'Politehnica' of Bucharest

Résultats de recherche: Le chapitre dans un livre, un rapport, une anthologie ou une collection › Contribution à une conférence › Revue par des pairs

Résumé

In this paper we have introduced a novel emotion recognition framework from raw speech signals. The system is based on ResNet architecture fed with spectrogram inputs. The CNN is further extended with a GhostVLAD feature aggregation layer that extracts a single, fixed size descriptor constructed at the level of the utterance. The system adopts a sentiment metric loss that integrates the relations between various classes of emotions. The experimental evaluation conducted on two publicly available databases: RAVDESS and CREMA-D validates the proposed methodology with average accuracy scores of 82% and 63%, respectively.

langue originale	Anglais
titre	2022 IEEE International Conference on Consumer Electronics, ICCE 2022
Editeur	Institute of Electrical and Electronics Engineers Inc.
ISBN (Electronique)	9781665441544
Les DOIs	https://doi.org/10.1109/ICCE53296.2022.9730534
état	Publié - 1 janv. 2022
Evénement	2022 IEEE International Conference on Consumer Electronics, ICCE 2022 - Virtual, Online, États-Unis Durée: 7 janv. 2022 → 9 janv. 2022

Série de publications

Nom	Digest of Technical Papers - IEEE International Conference on Consumer Electronics
Volume	2022-January
ISSN (imprimé)	0747-668X

Une conférence

Une conférence	2022 IEEE International Conference on Consumer Electronics, ICCE 2022
Pays/Territoire	États-Unis
La ville	Virtual, Online
période	7/01/22 → 9/01/22

Accès au document

10.1109/ICCE53296.2022.9730534

Contient cette citation

Mocanu, B., & Tapu, R. (2022). Emotion Recognition from Raw Speech Signals Using 2D CNN with Deep Metric Learning. Dans 2022 IEEE International Conference on Consumer Electronics, ICCE 2022 (Digest of Technical Papers - IEEE International Conference on Consumer Electronics; Vol 2022-January). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/ICCE53296.2022.9730534

@inproceedings{f0a3da751e834cbf97902e633ef97535,

title = "Emotion Recognition from Raw Speech Signals Using 2D CNN with Deep Metric Learning",

abstract = "In this paper we have introduced a novel emotion recognition framework from raw speech signals. The system is based on ResNet architecture fed with spectrogram inputs. The CNN is further extended with a GhostVLAD feature aggregation layer that extracts a single, fixed size descriptor constructed at the level of the utterance. The system adopts a sentiment metric loss that integrates the relations between various classes of emotions. The experimental evaluation conducted on two publicly available databases: RAVDESS and CREMA-D validates the proposed methodology with average accuracy scores of 82\% and 63\%, respectively.",

keywords = "GhostVLAD aggregation layer, multi-stage training, sentiment metric learning, speech emotion recognition",

author = "Bogdan Mocanu and Ruxandra Tapu",

note = "Publisher Copyright: {\textcopyright} 2022 IEEE.; 2022 IEEE International Conference on Consumer Electronics, ICCE 2022 ; Conference date: 07-01-2022 Through 09-01-2022",

year = "2022",

month = jan,

day = "1",

doi = "10.1109/ICCE53296.2022.9730534",

language = "English",

series = "Digest of Technical Papers - IEEE International Conference on Consumer Electronics",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

booktitle = "2022 IEEE International Conference on Consumer Electronics, ICCE 2022",

}

Mocanu, B & Tapu, R 2022, Emotion Recognition from Raw Speech Signals Using 2D CNN with Deep Metric Learning. Dans 2022 IEEE International Conference on Consumer Electronics, ICCE 2022. Digest of Technical Papers - IEEE International Conference on Consumer Electronics, VOL. 2022-January, Institute of Electrical and Electronics Engineers Inc., 2022 IEEE International Conference on Consumer Electronics, ICCE 2022, Virtual, Online, États-Unis, 7/01/22. https://doi.org/10.1109/ICCE53296.2022.9730534

Emotion Recognition from Raw Speech Signals Using 2D CNN with Deep Metric Learning. / Mocanu, Bogdan; Tapu, Ruxandra.
2022 IEEE International Conference on Consumer Electronics, ICCE 2022. Institute of Electrical and Electronics Engineers Inc., 2022. (Digest of Technical Papers - IEEE International Conference on Consumer Electronics; Vol 2022-January).

Résultats de recherche: Le chapitre dans un livre, un rapport, une anthologie ou une collection › Contribution à une conférence › Revue par des pairs

TY - GEN

T1 - Emotion Recognition from Raw Speech Signals Using 2D CNN with Deep Metric Learning

AU - Mocanu, Bogdan

AU - Tapu, Ruxandra

PY - 2022/1/1

Y1 - 2022/1/1

N2 - In this paper we have introduced a novel emotion recognition framework from raw speech signals. The system is based on ResNet architecture fed with spectrogram inputs. The CNN is further extended with a GhostVLAD feature aggregation layer that extracts a single, fixed size descriptor constructed at the level of the utterance. The system adopts a sentiment metric loss that integrates the relations between various classes of emotions. The experimental evaluation conducted on two publicly available databases: RAVDESS and CREMA-D validates the proposed methodology with average accuracy scores of 82% and 63%, respectively.

AB - In this paper we have introduced a novel emotion recognition framework from raw speech signals. The system is based on ResNet architecture fed with spectrogram inputs. The CNN is further extended with a GhostVLAD feature aggregation layer that extracts a single, fixed size descriptor constructed at the level of the utterance. The system adopts a sentiment metric loss that integrates the relations between various classes of emotions. The experimental evaluation conducted on two publicly available databases: RAVDESS and CREMA-D validates the proposed methodology with average accuracy scores of 82% and 63%, respectively.

KW - GhostVLAD aggregation layer

KW - multi-stage training

KW - sentiment metric learning

KW - speech emotion recognition

U2 - 10.1109/ICCE53296.2022.9730534

DO - 10.1109/ICCE53296.2022.9730534

M3 - Conference contribution

AN - SCOPUS:85127059517

T3 - Digest of Technical Papers - IEEE International Conference on Consumer Electronics

BT - 2022 IEEE International Conference on Consumer Electronics, ICCE 2022

PB - Institute of Electrical and Electronics Engineers Inc.

T2 - 2022 IEEE International Conference on Consumer Electronics, ICCE 2022

Y2 - 7 January 2022 through 9 January 2022

ER -

Emotion Recognition from Raw Speech Signals Using 2D CNN with Deep Metric Learning

Résumé

Série de publications

Une conférence

Accès au document

Empreinte digitale

Contient cette citation