Self-Supervised Learning for Few-Shot Bird Sound Classification
This work addresses the challenge of few-shot bird sound classification for bioacoustics researchers, but it is incremental as it applies existing self-supervised learning methods to a specific domain.
The paper tackled the problem of classifying bird sounds with limited labeled data by using self-supervised learning to learn representations from unlabeled audio, achieving improved generalization to new bird species in few-shot scenarios.
Self-supervised learning (SSL) in audio holds significant potential across various domains, particularly in situations where abundant, unlabeled data is readily available at no cost. This is pertinent in bioacoustics, where biologists routinely collect extensive sound datasets from the natural environment. In this study, we demonstrate that SSL is capable of acquiring meaningful representations of bird sounds from audio recordings without the need for annotations. Our experiments showcase that these learned representations exhibit the capacity to generalize to new bird species in few-shot learning (FSL) scenarios. Additionally, we show that selecting windows with high bird activation for self-supervised learning, using a pretrained audio neural network, significantly enhances the quality of the learned representations.