Homie0609/SoccerReplay-1988

Here is the dataset page of the dataset SoccerReplay-1988. Welcome all the soccer and AI to this page!! ⚽️⚽️🎉🎉 SoccerReplay-1988 is the largest soccer multi-modal dataset proposed in our work of Towards Universal Soccer Video Understanding in CVPR 2025. More information about this work could be found in our project page.

NDA FORM CLICK HERE!!! (You have to submit for requesting data)

If you want to access the dataset, to make sure these data WON'T BE USED FOR COMMERCIAL!! You have to fill out the NDA form, where you have to get agreement in this file. After that we will give you access to this huggingface dataset files (for textual part), and we will send you the link to the storage of the full-game videos of SoccerReplay-1988. The review process is often completed within a few more weeks.

Dataset Overview

The dataset includes 1988 soccer games from European major soccer leagues, with split of 1488/250/250 for train/valid/test sets.

SoccerReplay-1988 Dataset Overview

SoccerReplay-1988 Dataset Overview. Left: Data Curation Pipeline. The collected soccer video data are automatically processed for temporal alignment, event summarization, and anonymization by our curation pipeline.
Right: Statistics of Soccer Datasets. Our SoccerReplay-1988 significantly surpasses existing datasets in both scale and diversity. Here, # Anno. and # Com. refer to the number of event annotations and textual commentaries, respectively.

Event Labels and Commentaries in SoccerReplay-1988

Event Labels and Commentaries in SoccerReplay-1988. SoccerReplay-1988 dataset consists of full-match videos, event descriptions, and match-related information of 1,988 soccer matches.
Left: Distribution of Event Labels of 24 Classes in SoccerReplay-1988. In our dataset, we summarized 24 classes of event labels for soccer scenarios with second-level timestamps.
Right: Word Cloud of Commentaries. For each event in our dataset, we also provide a textual commentary to describe events with more details.

Many thanks for pioneering works from SoccerNet in AI Soccer. For any problems, feel free to contact jy_rao[at]sjtu[dot]edu[dot]cn or haoningwu3639[at]gmail[dot]com. Hope we can build a better AI for Sports community together, that is always our dreams and goals.