Re-Identification

SoccerNet Re-Identification

In this task, participants will have to re-identify soccer players across multiple camera viewpoints. This task is particularly useful for player tracking across multiple cameras. For instance click on a player and get all the frames where he appears. This will be a key component for building advanced automatic highlights solutions, such as customized videos focusing on a single player, or for developing better tools to support VAR referees.


Our task.

Given corresponding action and replay frames, and player bounding boxes in these frames, try to identify each player across the different camera views. Basically, this means linking the bounding boxes between these video frames from different viewpoints.

All of our classes.

The classes are not taken into account in the baseline or the evaluation, but the object to retrieve are among the following classes: {player team left, player team right, goalkeeper team left, goalkeeper team right, main referee, side referee, staff}.

Our data.

The SoccerNet Re-Identification (ReID) dataset is composed of 340.993 players thumbnails extracted from the SoccerNet videos at events annotated for the action spotting task, and images from their replays. The challenge set is composed of separate players thumbnails from different games.

Our Metric.

The retrieval-mAP is used as the main metric for this task.

For more details, check out our development kit on github

Our videos on Re-Identification

Soccer Player Tracking, Re-ID, Camera Calibration and Action Spotting - SoccerNet Challenges 2022

In this video, we present our new SoccerNet Challenges for CVPR 2022! We introduce the three tasks of Calibration, Re-identification and Tracking on soccer games, in partnership with EVS Broadcast Equipment, SportRadar and Baidu Research. We also reiterate our previous Action Spotting and Replay Grounding Challenges at the ActivityNet workshop.

How to cite this work ?

available soon