2026 BioDCASE Task 2 : Evaluation set

Le résumé fourni par la source

General description This repository contains the audio files of the evaluation set for the task 2 "Supervised Detection of Strongly-Labelled Whale Calls" of the 2026 BioDCASE challenge. Folder structure 2026_BioDCASE_evaluation_set.zip |___2026_BioDCASE_evaluation_set/ |____evaluation/ |____audio/ |____ddu2019/ |____*.wav |____ddu2021/ |____*.wav |____kerguelen2019/ |____*.wav |____kerguelen2020/ |____*.wav Dataset statistics Dataset Number of audio recordings Total audio duration (h) ddu2019 187 187 ddu2021 208 208 kerguelen2019 200 200 kerguelen2020 200 200 Submission In accordance with the general rules of the BioDCASE challenge, participants are reminded that no annotation data are provided. Instead, participants must submit their model outputs, which will then be evaluated and scored by the task organizers. On the task webpage, please refer to the section "Annotation"for further information on the annotation data (see also the development Zenodo repository) "Rules and submission" for information on how formatting your model outputs prior to submission Contact info Please, send any feedback or question to : Dorian CAZAU (dorian.cazau@ensta.fr) | Lucie JEAN-LABADYE (lucie.jean-labadye@sorbonne-universite.fr) | Cléa PARCERISAS (clea.parcerisas@vliz.be)

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.