AG-VPReID 2025: Aerial-Ground Video-based Person Re-identification Challenge Results

TL;DR

The paper introduces the first large-scale video-based aerial-ground person re-identification challenge, presenting a new dataset and showcasing advanced solutions that significantly improve re-identification accuracy across viewpoints.

Contribution

It presents the AG-VPReID 2025 Challenge, a novel large-scale dataset and benchmark for aerial-ground video person re-identification, along with state-of-the-art solution approaches.

Findings

X-TFCLIP achieved 72.28% Rank-1 accuracy in aerial-ground ReID

The dataset contains over 3.7 million frames from UAVs and cameras

Multiple innovative solutions outperformed existing baselines

Abstract

Person re-identification (ReID) across aerial and ground vantage points has become crucial for large-scale surveillance and public safety applications. Although significant progress has been made in ground-only scenarios, bridging the aerial-ground domain gap remains a formidable challenge due to extreme viewpoint differences, scale variations, and occlusions. Building upon the achievements of the AG-ReID 2023 Challenge, this paper introduces the AG-VPReID 2025 Challenge - the first large-scale video-based competition focused on high-altitude (80-120m) aerial-ground ReID. Constructed on the new AG-VPReID dataset with 3,027 identities, over 13,500 tracklets, and approximately 3.7 million frames captured from UAVs, CCTV, and wearable cameras, the challenge featured four international teams. These teams developed solutions ranging from multi-stream architectures to transformer-based temporal reasoning and physics-informed modeling. The leading approach, X-TFCLIP from UAM, attained 72.28% Rank-1 accuracy in the aerial-to-ground ReID setting and 70.77% in the ground-to-aerial ReID setting, surpassing existing baselines while highlighting the dataset's complexity. For additional details, please refer to the official website at https://agvpreid25.github.io.