Semi-supervised classification of bird vocalizations

TL;DR

This paper introduces a semi-supervised bird call detector that can identify overlapping calls and requires minimal labeled data, outperforming existing methods on diverse soundscape recordings.

Contribution

It presents a novel semi-supervised classifier capable of detecting overlapping bird calls with few labeled samples, improving over state-of-the-art methods in complex soundscapes.

Findings

Achieves a mean F0.5 score of 0.701 on 315 classes from 110 species.

Outperforms BirdNET with fewer labeled training samples.

Effective in continuous soundscape data despite false positive challenges.

Abstract

Changes in bird populations can indicate broader changes in ecosystems, making birds one of the most important animal groups to monitor. Combining machine learning and passive acoustics enables continuous monitoring over extended periods without direct human involvement. However, most existing techniques require extensive expert-labeled datasets for training and cannot easily detect time-overlapping calls in busy soundscapes. We propose a semi-supervised acoustic bird detector designed to allow both the detection of time-overlapping calls (when separated in frequency) and the use of few labeled training samples. The classifier is trained and evaluated on a combination of community-recorded open-source data and long-duration soundscape recordings from Singapore. It achieves a mean F0.5 score of 0.701 across 315 classes from 110 bird species on a hold-out test set, with an average of 11 labeled training samples per class. It outperforms the state-of-the-art BirdNET classifier on a test set of 103 bird species despite significantly fewer labeled training samples. The detector is further tested on 144 microphone-hours of continuous soundscape data. The rich soundscape in Singapore makes suppression of false positives a challenge on raw, continuous data streams. Nevertheless, we demonstrate that achieving high precision in such environments with minimal labeled training data is possible.