GTA: Global Tracklet Association for Multi-Object Tracking in Sports
This addresses challenges like player re-identification and ID switches in sports tracking, offering a plug-and-play refinement tool for existing trackers, though it is incremental as it builds on prior methods.
The paper tackles the problem of multi-object tracking in sports scenarios by proposing an appearance-based global tracklet association algorithm that splits tracklets with multiple identities and connects those from the same identity, achieving a new state-of-the-art HOTA score of 81.04% on the SportsMOT dataset and improving scores from 79.41% to 83.11% on the SoccerNet dataset.
Multi-object tracking in sports scenarios has become one of the focal points in computer vision, experiencing significant advancements through the integration of deep learning techniques. Despite these breakthroughs, challenges remain, such as accurately re-identifying players upon re-entry into the scene and minimizing ID switches. In this paper, we propose an appearance-based global tracklet association algorithm designed to enhance tracking performance by splitting tracklets containing multiple identities and connecting tracklets seemingly from the same identity. This method can serve as a plug-and-play refinement tool for any multi-object tracker to further boost their performance. The proposed method achieved a new state-of-the-art performance on the SportsMOT dataset with HOTA score of 81.04%. Similarly, on the SoccerNet dataset, our method enhanced multiple trackers' performance, consistently increasing the HOTA score from 79.41% to 83.11%. These significant and consistent improvements across different trackers and datasets underscore our proposed method's potential impact on the application of sports player tracking. We open-source our project codebase at https://github.com/sjc042/gta-link.git.