Son Année : 2025

Audiocarnet - Deep Representation Learning from Unlabeled Bioacoustic Data

Résumé

Contrastive learning requires creating distinct views of the same input while preserving important details. Standard audio augmentations, such as random resized crops, can remove species-specific characteristics. We propose mixing vocalizations as a domain-agnostic data augmentation, which preserves the unique features of the species of interest while forming a distinct view. This simple strategy allows contrastive learning to capture species-specific features in bird vocalizations from unlabeled data.

Fichier principal
Vignette du fichier
moummad_bioacoustic_data_augmentation.wav (11.57 Mo) Télécharger le fichier
Origine Fichiers produits par l'(les) auteur(s)
Licence

Dates et versions

hal-05342505 , version 1 (02-11-2025)

Licence

Identifiants

  • HAL Id : hal-05342505 , version 1

Citer

Ilyass Moummad. Audiocarnet - Deep Representation Learning from Unlabeled Bioacoustic Data. 2025. ⟨hal-05342505⟩
266 Consultations
44 Téléchargements

Partager

  • More