EventAnchor: Reducing Human Interactions in Event Annotation of Racket Sports Videos

TL;DR

EventAnchor is a framework that leverages computer vision and machine learning to simplify and improve the accuracy of annotating racket sports videos, reducing human effort and domain expertise needed.

Contribution

It introduces a novel interactive annotation framework that integrates computer vision models to assist users in identifying key events in racket sports videos.

Findings

Significant improvement in annotation accuracy and efficiency.

Effective support for both simple and complex annotation tasks.

Enhanced user performance with domain knowledge requirements.

Abstract

The popularity of racket sports (e.g., tennis and table tennis) leads to high demands for data analysis, such as notational analysis, on player performance. While sports videos offer many benefits for such analysis, retrieving accurate information from sports videos could be challenging. In this paper, we propose EventAnchor, a data analysis framework to facilitate interactive annotation of racket sports video with the support of computer vision algorithms. Our approach uses machine learning models in computer vision to help users acquire essential events from videos (e.g., serve, the ball bouncing on the court) and offers users a set of interactive tools for data annotation. An evaluation study on a table tennis annotation system built on this framework shows significant improvement of user performances in simple annotation tasks on objects of interest and complex annotation tasks requiring domain knowledge.