On the Utility of Speech and Audio Foundation Models for Marmoset Call Analysis

TL;DR

This study evaluates the effectiveness of speech and audio foundation models in classifying marmoset calls, finding that higher bandwidth models and pre-training on speech or general audio improve classification accuracy over traditional spectral methods.

Contribution

It systematically compares speech and general audio pre-trained models across different bandwidths for marmoset call classification, highlighting their utility and limitations.

Findings

Higher bandwidth models improve classification performance.

Pre-training on speech and general audio yields similar results.

Models outperform traditional spectral baseline methods.

Abstract

Marmoset monkeys encode vital information in their calls and serve as a surrogate model for neuro-biologists to understand the evolutionary origins of human vocal communication. Traditionally analyzed with signal processing-based features, recent approaches have utilized self-supervised models pre-trained on human speech for feature extraction, capitalizing on their ability to learn a signal's intrinsic structure independently of its acoustic domain. However, the utility of such foundation models remains unclear for marmoset call analysis in terms of multi-class classification, bandwidth, and pre-training domain. This study assesses feature representations derived from speech and general audio domains, across pre-training bandwidths of 4, 8, and 16 kHz for marmoset call-type and caller classification tasks. Results show that models with higher bandwidth improve performance, and pre-training on speech or general audio yields comparable results, improving over a spectral baseline.