SoccerDB: A Large-Scale Database for Comprehensive Video Understanding

TL;DR

SoccerDB is a large-scale, comprehensive soccer video database designed to facilitate research in various video understanding tasks, including detection, recognition, and localization, by providing extensive annotated data.

Contribution

The paper introduces SoccerDB, the largest sports video database with detailed annotations, enabling multi-task research and advancing comprehensive video understanding.

Findings

State-of-the-art baseline performances on individual tasks.

Joint task consideration improves overall understanding.

Dataset and code publicly available for research.

Abstract

Soccer videos can serve as a perfect research object for video understanding because soccer games are played under well-defined rules while complex and intriguing enough for researchers to study. In this paper, we propose a new soccer video database named SoccerDB, comprising 171,191 video segments from 346 high-quality soccer games. The database contains 702,096 bounding boxes, 37,709 essential event labels with time boundary and 17,115 highlight annotations for object detection, action recognition, temporal action localization, and highlight detection tasks. To our knowledge, it is the largest database for comprehensive sports video understanding on various aspects. We further survey a collection of strong baselines on SoccerDB, which have demonstrated state-of-the-art performances on independent tasks. Our evaluation suggests that we can benefit significantly when jointly considering the inner correlations among those tasks. We believe the release of SoccerDB will tremendously advance researches around comprehensive video understanding. {\itshape Our dataset and code published on https://github.com/newsdata/SoccerDB.}