Great Ape Detection in Challenging Jungle Camera Trap Footage via Attention-Based Spatial and Temporal Feature Blending

TL;DR

This paper introduces a novel multi-frame video detection framework with attention-based feature blending for identifying great apes in challenging jungle camera trap footage, significantly improving detection robustness.

Contribution

The paper presents the first multi-frame detection method incorporating self-attention for spatial and temporal feature blending in wildlife monitoring.

Findings

Outperforms frame-based detectors in challenging conditions

Achieves high robustness on real-world camera trap data

Demonstrates effectiveness on large-scale annotated datasets

Abstract

We propose the first multi-frame video object detection framework trained to detect great apes. It is applicable to challenging camera trap footage in complex jungle environments and extends a traditional feature pyramid architecture by adding self-attention driven feature blending in both the spatial as well as the temporal domain. We demonstrate that this extension can detect distinctive species appearance and motion signatures despite significant partial occlusion. We evaluate the framework using 500 camera trap videos of great apes from the Pan African Programme containing 180K frames, which we manually annotated with accurate per-frame animal bounding boxes. These clips contain significant partial occlusions, challenging lighting, dynamic backgrounds, and natural camouflage effects. We show that our approach performs highly robustly and significantly outperforms frame-based detectors. We also perform detailed ablation studies and validation on the full ILSVRC 2015 VID data corpus to demonstrate wider applicability at adequate performance levels. We conclude that the framework is ready to assist human camera trap inspection efforts. We publish code, weights, and ground truth annotations with this paper.