Few-shot bioacoustic event detection at the DCASE 2022 challenge

TL;DR

This paper reviews the second edition of a challenge on few-shot bioacoustic event detection, highlighting advances in methods like prototypical networks and the remaining challenges in the field.

Contribution

It provides a comprehensive overview of the challenge, datasets, baselines, and state-of-the-art methods, demonstrating significant progress and identifying ongoing difficulties.

Findings

13 out of 15 teams outperformed baselines

Highest F-score reached 60%

Methods used include prototypical networks and transductive learning

Abstract

Few-shot sound event detection is the task of detecting sound events, despite having only a few labelled examples of the class of interest. This framework is particularly useful in bioacoustics, where often there is a need to annotate very long recordings but the expert annotator time is limited. This paper presents an overview of the second edition of the few-shot bioacoustic sound event detection task included in the DCASE 2022 challenge. A detailed description of the task objectives, dataset, and baselines is presented, together with the main results obtained and characteristics of the submitted systems. This task received submissions from 15 different teams from which 13 scored higher than the baselines. The highest F-score was of 60% on the evaluation set, which leads to a huge improvement over last year's edition. Highly-performing methods made use of prototypical networks, transductive learning, and addressed the variable length of events from all target classes. Furthermore, by analysing results on each of the subsets we can identify the main difficulties that the systems face, and conclude that few-show bioacoustic sound event detection remains an open challenge.