AI system to
predict teams resuming matches from ball out footage

Yuma Hirose

Shizuoka University

Makoto Okabe

Shizuoka University

AI Predicts Resumed Teams from 5-second Videos Including Ball-Out Scenes, Reducing Referee Burden and Improving Consistency of Decisions

Abstract

This research will develop an AI-based system to automate the resumption team decision when the ball is out in a soccer match.Currently, VAR and goal line technology assist in important decisions, but do not support ball-out decisions.This system will use match video as input, and AI will predict the resuming team based on player movement and uniform color.This system is expected to realize quick and consistent judgments and contribute to reducing the burden on referees and improving the consistency of judgments.

Paper

Eurographics 2005

Video

BibTeX citation

    @Misc{minami.takaaki.19_bachelor,
        author =    {Yuma Hirose},
        title =     {AI system to predict teams resuming matches from ball out footage},
        howpublished = {Bachelor Thesis},
        month =     2,
        year =      2025,
    }