Adaptive Learning via a Negative Selection Strategy for Few-Shot Bioacoustic Event Detection

TL;DR

This paper introduces an adaptive learning framework with a negative selection strategy to improve few-shot bioacoustic event detection, addressing negative prototype construction and vocalization duration variability, resulting in significant performance gains.

Contribution

It proposes a novel adaptive learning loss and negative selection strategy to enhance ProtoNet for few-shot bioacoustic detection, handling negative prototype construction and variable vocalization durations.

Findings

Achieved an F-measure of 0.703 on DCASE 2023 TASK5 dataset.

Improved performance by 12.84% over baseline methods.

Effectively addresses negative prototype and duration variability issues.

Abstract

Although the Prototypical Network (ProtoNet) has demonstrated effectiveness in few-shot biological event detection, two persistent issues remain. Firstly, there is difficulty in constructing a representative negative prototype due to the absence of explicitly annotated negative samples. Secondly, the durations of the target biological vocalisations vary across tasks, making it challenging for the model to consistently yield optimal results across all tasks. To address these issues, we propose a novel adaptive learning framework with an adaptive learning loss to guide classifier updates. Additionally, we propose a negative selection strategy to construct a more representative negative prototype for ProtoNet. All experiments ware performed on the DCASE 2023 TASK5 few-shot bioacoustic event detection dataset. The results show that our proposed method achieves an F-measure of 0.703, an improvement of 12.84%.