GTA: Global Tracklet Association for Multi-Object Tracking in Sports

TL;DR

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 is proposed.

Abstract

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.

Figures (8)

• Figure 1: Our proposed Global Tracklet Association (GTA) method significantly boosts the HOTA and IDF1 score of existing trackers, such as SORT, ByteTrack, and Deep-EIoU, on sports tracking datasets, including SportsMOT and SoccerNet.
• Figure 2: Examples of different players on the same teams, highlighting the challenge of distinguishing between players with similar appearances in sports tracking.
• Figure 3: An example of a mix-up error in a single tracklet. The tracklet output by the online tracking system contains three different identities, represented by purple, green, and red points. The figure illustrates the tracklet's features extracted by a ReID model and clustered using the DBSCAN clustering algorithm.
• Figure 4: An example of a cut-off error, where a player's tracklet is fragmented into four separate segments due to the player exiting and re-entering the camera view multiple times throughout the video sequence.
• Figure 5: Illustration of tracklet splitter.