GTATrack: Winner Solution to SoccerTrack 2025 with Deep-EIoU and Global Tracklet Association

2025-10-27

Proceedings of the 8th International ACM Workshop on Multimedia Content Analysis in Sports

AI Summary

This work addresses the challenging problem of multi-object tracking in soccer scenes captured by static fisheye cameras, where irregular motion, appearance similarity, frequent occlusions, and geometric distortion severely degrade tracking performance. To tackle these issues, the authors propose GTATrack, a hierarchical tracking framework that integrates Deep-EIoU for motion-agnostic online association and introduces Global Trajectory Association (GTA) for trajectory-level refinement. A pseudo-labeling strategy is further employed to enhance the recall of small-object detection. By synergistically combining local matching with global reasoning, the method effectively mitigates identity switches, occlusion-induced failures, and trajectory fragmentation. Evaluated on the SoccerTrack 2025 Challenge, GTATrack achieves state-of-the-art performance with an HOTA score of 0.60 and only 982 false positives, significantly outperforming existing approaches in fisheye-based soccer tracking scenarios.

Technology Category

• Computer Vision: Motion & Tracking
• Intelligent Robots: State Estimation
• Knowledge Representation and Reasoning: Geometric, Spatial, and Temporal Reasoning

Application Category

• Security and Privacy: Tracking, profiling, and countermeasures against them
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs

Abstract

Multi-object tracking (MOT) in sports is highly challenging due to irregular player motion, uniform appearances, and frequent occlusions. These difficulties are further exacerbated by the geometric distortion and extreme scale variation introduced by static fisheye cameras. In this work, we present GTATrack, a hierarchical tracking framework that win first place in the SoccerTrack Challenge 2025. GTATrack integrates two core components: Deep Expansion IoU (Deep-EIoU) for motion-agnostic online association and Global Tracklet Association (GTA) for trajectory-level refinement. This two-stage design enables both robust short-term matching and long-term identity consistency. Additionally, a pseudo-labeling strategy is used to boost detector recall on small and distorted targets. The synergy between local association and global reasoning effectively addresses identity switches, occlusions, and tracking fragmentation. Our method achieved a winning HOTA score of 0.60 and significantly reduced false positives to 982, demonstrating state-of-the-art accuracy in fisheye-based soccer tracking.

Problem

• multi-object tracking
• fisheye camera
• occlusions
• identity switches
• scale variation

Innovation

• Deep-EIoU
• Global Tracklet Association
• multi-object tracking
• fisheye camera
• pseudo-labeling