Deep Learning for Leopard Individual Identification: An Adaptive Angular Margin Approach

TL;DR

This paper presents a deep learning framework with an adaptive angular margin for identifying individual leopards from camera trap images, improving accuracy over baseline methods and aiding wildlife monitoring.

Contribution

It introduces a novel adaptive angular margin method within a deep learning architecture and a preprocessing pipeline combining RGB and edge detection for wildlife identification.

Findings

Achieved a Dynamic Top-5 Average Precision of 0.8814

Reached a Top-5 Rank Match Detection of 0.9533

Outperformed the Triplet Network baseline

Abstract

Accurate identification of individual leopards across camera trap images is critical for population monitoring and ecological studies. This paper introduces a deep learning framework to distinguish between individual leopards based on their unique spot patterns. This approach employs a novel adaptive angular margin method in the form of a modified CosFace architecture. In addition, I propose a preprocessing pipeline that combines RGB channels with an edge detection channel to underscore the critical features learned by the model. This approach significantly outperforms the Triplet Network baseline, achieving a Dynamic Top-5 Average Precision of 0.8814 and a Top-5 Rank Match Detection of 0.9533, demonstrating its potential for open-set learning in wildlife identification. While not surpassing the performance of the SIFT-based Hotspotter algorithm, this method represents a substantial advancement in applying deep learning to patterned wildlife identification. This research contributes to the field of computer vision and provides a valuable tool for biologists aiming to study and protect leopard populations. It also serves as a stepping stone for applying the power of deep learning in Capture-Recapture studies for other patterned species.