An Ensemble Deep-Learning System for Few-Shot Bioacoustic Event Detection

Jianqian Zhang; Miao Liu; Jing Wang; Chenguang Hu; Jiawei Peng; Kaige Li

doi:10.1109/ACAIT60137.2023.10528451

An Ensemble Deep-Learning System for Few-Shot Bioacoustic Event Detection

Jianqian Zhang, Miao Liu, Jing Wang^*, Chenguang Hu, Jiawei Peng, Kaige Li

^*Corresponding author for this work

School of Information and Electronics

Beijing Institute of Technology

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

Abstract

Human are good at identifying new objects from only a few number of samples. Inspired by the rapid learning ability of humans, few-shot bioacoustic event detection task based on machine learning method has attracted more attention lately. In this paper, we formulated an ensemble deep-learning system to address the task, incorporating an enhanced prototypical network and transductive inference methodology. We have submitted our models to Task 5 of the Detection and Classification of Acoustic Scenes and Events 2022 (DCASE2022) challenge, and got F-measure of 64.8% on the validation set as our best score. In the final competition, our system got F-measure of 44.3% on the evaluation set.

Original language	English
Title of host publication	Proceedings of 2023 7th Asian Conference on Artificial Intelligence Technology, ACAIT 2023
Publisher	Institute of Electrical and Electronics Engineers Inc.
Pages	192-197
Number of pages	6
ISBN (Electronic)	9798350359145
DOIs	https://doi.org/10.1109/ACAIT60137.2023.10528451
Publication status	Published - 2023
Event	7th Asian Conference on Artificial Intelligence Technology, ACAIT 2023 - Quzhou, China Duration: 10 Nov 2023 → 12 Nov 2023

Publication series

Name	Proceedings of 2023 7th Asian Conference on Artificial Intelligence Technology, ACAIT 2023

Conference

Conference	7th Asian Conference on Artificial Intelligence Technology, ACAIT 2023
Country/Territory	China
City	Quzhou
Period	10/11/23 → 12/11/23

Keywords

bioacoustic event detection
DCASE
few-shot task
prototypical networks
transductive inference

Access to Document

10.1109/ACAIT60137.2023.10528451

Cite this

Zhang, J., Liu, M., Wang, J., Hu, C., Peng, J., & Li, K. (2023). An Ensemble Deep-Learning System for Few-Shot Bioacoustic Event Detection. In Proceedings of 2023 7th Asian Conference on Artificial Intelligence Technology, ACAIT 2023 (pp. 192-197). (Proceedings of 2023 7th Asian Conference on Artificial Intelligence Technology, ACAIT 2023). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/ACAIT60137.2023.10528451

Zhang, Jianqian ; Liu, Miao ; Wang, Jing et al. / An Ensemble Deep-Learning System for Few-Shot Bioacoustic Event Detection. Proceedings of 2023 7th Asian Conference on Artificial Intelligence Technology, ACAIT 2023. Institute of Electrical and Electronics Engineers Inc., 2023. pp. 192-197 (Proceedings of 2023 7th Asian Conference on Artificial Intelligence Technology, ACAIT 2023).

@inproceedings{c1fabdbdd9d745aea75567c35268186c,

title = "An Ensemble Deep-Learning System for Few-Shot Bioacoustic Event Detection",

abstract = "Human are good at identifying new objects from only a few number of samples. Inspired by the rapid learning ability of humans, few-shot bioacoustic event detection task based on machine learning method has attracted more attention lately. In this paper, we formulated an ensemble deep-learning system to address the task, incorporating an enhanced prototypical network and transductive inference methodology. We have submitted our models to Task 5 of the Detection and Classification of Acoustic Scenes and Events 2022 (DCASE2022) challenge, and got F-measure of 64.8% on the validation set as our best score. In the final competition, our system got F-measure of 44.3% on the evaluation set.",

keywords = "bioacoustic event detection, DCASE, few-shot task, prototypical networks, transductive inference",

author = "Jianqian Zhang and Miao Liu and Jing Wang and Chenguang Hu and Jiawei Peng and Kaige Li",

note = "Publisher Copyright: {\textcopyright} 2023 IEEE.; 7th Asian Conference on Artificial Intelligence Technology, ACAIT 2023 ; Conference date: 10-11-2023 Through 12-11-2023",

year = "2023",

doi = "10.1109/ACAIT60137.2023.10528451",

language = "English",

series = "Proceedings of 2023 7th Asian Conference on Artificial Intelligence Technology, ACAIT 2023",

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

pages = "192--197",

booktitle = "Proceedings of 2023 7th Asian Conference on Artificial Intelligence Technology, ACAIT 2023",

address = "United States",

}

Zhang, J, Liu, M, Wang, J, Hu, C, Peng, J & Li, K 2023, An Ensemble Deep-Learning System for Few-Shot Bioacoustic Event Detection. in Proceedings of 2023 7th Asian Conference on Artificial Intelligence Technology, ACAIT 2023. Proceedings of 2023 7th Asian Conference on Artificial Intelligence Technology, ACAIT 2023, Institute of Electrical and Electronics Engineers Inc., pp. 192-197, 7th Asian Conference on Artificial Intelligence Technology, ACAIT 2023, Quzhou, China, 10/11/23. https://doi.org/10.1109/ACAIT60137.2023.10528451

An Ensemble Deep-Learning System for Few-Shot Bioacoustic Event Detection. / Zhang, Jianqian; Liu, Miao; Wang, Jing et al.
Proceedings of 2023 7th Asian Conference on Artificial Intelligence Technology, ACAIT 2023. Institute of Electrical and Electronics Engineers Inc., 2023. p. 192-197 (Proceedings of 2023 7th Asian Conference on Artificial Intelligence Technology, ACAIT 2023).

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

TY - GEN

T1 - An Ensemble Deep-Learning System for Few-Shot Bioacoustic Event Detection

AU - Zhang, Jianqian

AU - Liu, Miao

AU - Wang, Jing

AU - Hu, Chenguang

AU - Peng, Jiawei

AU - Li, Kaige

PY - 2023

Y1 - 2023

N2 - Human are good at identifying new objects from only a few number of samples. Inspired by the rapid learning ability of humans, few-shot bioacoustic event detection task based on machine learning method has attracted more attention lately. In this paper, we formulated an ensemble deep-learning system to address the task, incorporating an enhanced prototypical network and transductive inference methodology. We have submitted our models to Task 5 of the Detection and Classification of Acoustic Scenes and Events 2022 (DCASE2022) challenge, and got F-measure of 64.8% on the validation set as our best score. In the final competition, our system got F-measure of 44.3% on the evaluation set.

AB - Human are good at identifying new objects from only a few number of samples. Inspired by the rapid learning ability of humans, few-shot bioacoustic event detection task based on machine learning method has attracted more attention lately. In this paper, we formulated an ensemble deep-learning system to address the task, incorporating an enhanced prototypical network and transductive inference methodology. We have submitted our models to Task 5 of the Detection and Classification of Acoustic Scenes and Events 2022 (DCASE2022) challenge, and got F-measure of 64.8% on the validation set as our best score. In the final competition, our system got F-measure of 44.3% on the evaluation set.

KW - bioacoustic event detection

KW - DCASE

KW - few-shot task

KW - prototypical networks

KW - transductive inference

UR - http://www.scopus.com/inward/record.url?scp=85194133265&partnerID=8YFLogxK

U2 - 10.1109/ACAIT60137.2023.10528451

DO - 10.1109/ACAIT60137.2023.10528451

M3 - Conference contribution

AN - SCOPUS:85194133265

T3 - Proceedings of 2023 7th Asian Conference on Artificial Intelligence Technology, ACAIT 2023

SP - 192

EP - 197

BT - Proceedings of 2023 7th Asian Conference on Artificial Intelligence Technology, ACAIT 2023

PB - Institute of Electrical and Electronics Engineers Inc.

T2 - 7th Asian Conference on Artificial Intelligence Technology, ACAIT 2023

Y2 - 10 November 2023 through 12 November 2023

ER -

Zhang J, Liu M, Wang J, Hu C, Peng J, Li K. An Ensemble Deep-Learning System for Few-Shot Bioacoustic Event Detection. In Proceedings of 2023 7th Asian Conference on Artificial Intelligence Technology, ACAIT 2023. Institute of Electrical and Electronics Engineers Inc. 2023. p. 192-197. (Proceedings of 2023 7th Asian Conference on Artificial Intelligence Technology, ACAIT 2023). doi: 10.1109/ACAIT60137.2023.10528451

An Ensemble Deep-Learning System for Few-Shot Bioacoustic Event Detection

Abstract

Publication series

Conference

Keywords

Access to Document

Other files and links

Fingerprint

Cite this