Few-shot Bioacoustic Event Detection with Machine Learning Methods

TL;DR

This paper explores a novel machine learning approach for few-shot bioacoustic event detection, successfully classifying rare animal sounds with limited training data, and compares different models in a challenging real-world setting.

Contribution

The study introduces a machine learning-based method for few-shot bioacoustic detection, diverging from common deep learning approaches, and evaluates multiple models on real field recordings.

Findings

Logistic regression outperformed linear regression and template matching.

All tested methods over-predicted the number of events.

The approach demonstrates potential for rare sound classification with limited data.

Abstract

Few-shot learning is a type of classification through which predictions are made based on a limited number of samples for each class. This type of classification is sometimes referred to as a meta-learning problem, in which the model learns how to learn to identify rare cases. We seek to extract information from five exemplar vocalisations of mammals or birds and detect and classify these sounds in field recordings [2]. This task was provided in the Detection and Classification of Acoustic Scenes and Events (DCASE) Challenge of 2021. Rather than utilize deep learning, as is most commonly done, we formulated a novel solution using only machine learning methods. Various models were tested, and it was found that logistic regression outperformed both linear regression and template matching. However, all of these methods over-predicted the number of events in the field recordings.