Self-Supervised Real-Time Tracking of Military Vehicles in Low-FPS UAV Footage

TL;DR

This paper introduces a robust real-time multi-object tracking method for low-FPS UAV footage of military vehicles, leveraging scene context and single-frame annotations to improve accuracy under challenging conditions.

Contribution

The paper proposes a novel instance association learning approach that uses scene global features and single-frame annotations to enhance low-FPS multi-object tracking in combat scenarios.

Findings

Effective tracking despite low resolution and frame rate

High association accuracy with scene context integration

Robustness to image degradation and distractors

Abstract

Multi-object tracking (MOT) aims to maintain consistent identities of objects across video frames. Associating objects in low-frame-rate videos captured by moving unmanned aerial vehicles (UAVs) in actual combat scenarios is complex due to rapid changes in object appearance and position within the frame. The task becomes even more challenging due to image degradation caused by cloud video streaming and compression algorithms. We present how instance association learning from single-frame annotations can overcome these challenges. We show that global features of the scene provide crucial context for low-FPS instance association, allowing our solution to be robust to distractors and gaps in detections. We also demonstrate that such a tracking approach maintains high association quality even when reducing the input image resolution and latent representation size for faster inference. Finally, we present a benchmark dataset of annotated military vehicles collected from publicly available data sources. This paper was initially presented at the NATO Science and Technology Organization Symposium (ICMCIS) organized by the Information Systems Technology (IST)Scientific and Technical Committee, IST-209-RSY - the ICMCIS, held in Oeiras, Portugal, 13-14 May 2025.