Deep embedded clustering of coral reef bioacoustics

TL;DR

This paper applies deep embedded clustering to coral reef soundscape signals, successfully distinguishing fish calls from whale songs and outperforming traditional clustering methods in accuracy.

Contribution

It introduces the use of deep embedded clustering for bioacoustic signal classification in coral reefs, demonstrating improved accuracy over conventional methods.

Findings

DEC achieved 77.5% accuracy on a small dataset.

Both GMM and DEC effectively identified fish and whale clusters.

Conventional clustering struggled with overlapping signals.

Abstract

Deep clustering was applied to unlabeled, automatically detected signals in a coral reef soundscape to distinguish fish pulse calls from segments of whale song. Deep embedded clustering (DEC) learned latent features and formed classification clusters using fixed-length power spectrograms of the signals. Handpicked spectral and temporal features were also extracted and clustered with Gaussian mixture models (GMM) and conventional clustering. DEC, GMM, and conventional clustering were tested on simulated datasets of fish pulse calls (fish) and whale song units (whale) with randomized bandwidth, duration, and SNR. Both GMM and DEC achieved high accuracy and identified clusters with fish, whale, and overlapping fish and whale signals. Conventional clustering methods had low accuracy in scenarios with unequal-sized clusters or overlapping signals. Fish and whale signals recorded near Hawaii in February-March 2020 were clustered with DEC, GMM, and conventional clustering. DEC features demonstrated the highest accuracy of 77.5% on a small, manually labeled dataset for classifying signals into fish and whale clusters.