Triple-stream Deep Metric Learning of Great Ape Behavioural Actions

TL;DR

This paper introduces a novel triple-stream deep metric learning system for recognizing great ape behaviors in wild camera trap videos, significantly improving accuracy and providing new insights into ape behavior spaces.

Contribution

The paper presents the first metric learning approach with a DensePose-C segmentation stream for great ape behavior recognition, achieving notable accuracy improvements.

Findings

~12% top-1 accuracy improvement over previous methods

~23% increase in per-class accuracy with long-tail recognition techniques

First data-driven visualization of ape behavioral action spaces

Abstract

We propose the first metric learning system for the recognition of great ape behavioural actions. Our proposed triple stream embedding architecture works on camera trap videos taken directly in the wild and demonstrates that the utilisation of an explicit DensePose-C chimpanzee body part segmentation stream effectively complements traditional RGB appearance and optical flow streams. We evaluate system variants with different feature fusion techniques and long-tail recognition approaches. Results and ablations show performance improvements of ~12% in top-1 accuracy over previous results achieved on the PanAf-500 dataset containing 180,000 manually annotated frames across nine behavioural actions. Furthermore, we provide a qualitative analysis of our findings and augment the metric learning system with long-tail recognition techniques showing that average per class accuracy -- critical in the domain -- can be improved by ~23% compared to the literature on that dataset. Finally, since our embedding spaces are constructed as metric, we provide first data-driven visualisations of the great ape behavioural action spaces revealing emerging geometry and topology. We hope that the work sparks further interest in this vital application area of computer vision for the benefit of endangered great apes.