Complex data labeling with deep learning methods: Lessons from fisheries acoustics - Archive ouverte HAL Accéder directement au contenu
Article Dans Une Revue ISA Transactions Année : 2021

Complex data labeling with deep learning methods: Lessons from fisheries acoustics

Résumé

Quantitative and qualitative analysis of acoustic backscattered signals from the seabed bottom to the sea surface is used worldwide for fish stocks assessment and marine ecosystem monitoring. Huge amounts of raw data are collected yet require tedious expert labeling. This paper focuses on a case study where the ground truth labels are non-obvious: echograms labeling, which is time-consuming and critical for the quality of fisheries and ecological analysis. We investigate how these tasks can benefit from supervised learning algorithms and demonstrate that convolutional neural networks trained with non-stationary datasets can be used to stress parts of a new dataset needing human expert correction. Further development of this approach paves the way toward a standardization of the labeling process in fisheries acoustics and is a good case study for non-obvious data labeling processes. (C) 2020 ISA. Published by Elsevier Ltd. All rights reserved.
Fichier principal
Vignette du fichier
S001905782030392X.pdf (1.31 Mo) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)

Dates et versions

hal-03246778 , version 1 (09-03-2023)

Licence

Paternité - Pas d'utilisation commerciale

Identifiants

Citer

Jean-Michel A. Sarr, Timothee Brochier, P. Brehmer, Y. Perrot, A. Bah, et al.. Complex data labeling with deep learning methods: Lessons from fisheries acoustics. ISA Transactions, 2021, 109, pp.113--125. ⟨10.1016/j.isatra.2020.09.018⟩. ⟨hal-03246778⟩
64 Consultations
52 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More