aceFaster, more accurate? A feasibility study on replacing human judges with artificial intelligence in video review for the Paris Olympics Taekwondo competition
Introduction: This study explores the potential of artificial intelligence (AI) to enhance the accuracy and efficiency of video review systems in Taekwondo, addressing limitations in current human-based judgment processes during competitions. Methods: A total of 241 video review cases from the 2024 Paris Olympic Taekwondo competition were analyzed. AI-based judgments were generated using ChatGPT-4.5 and OpenPose deep learning models. The AI-generated penalty decisions were statistically compared to those made by…
In a study of Paris 2024 taekwondo video reviews, ChatGPT-4.5 with OpenPose agreed almost perfectly with international referees (kappa 0.897) and cut average review time by about 81%.
The same AI disagreed with referees in nine reviews, mostly minimal-contact head strikes or blocked views, and an expert panel classed two as outright AI errors.
The authors say AI cannot yet replace human referees: it struggled with minimal contact, occlusion and ambiguous calls, and the work covers one event, one sport and frame-by-frame analysis rather than real-time, multi-angle review.
Sources
- Peer-reviewedFrontiers in Sports and Active Living2025-08-18
What do you make of this one?
Truvace Impact Record TRV-2026-2398, v1: “Faster, more accurate? A feasibility study on replacing human judges with artificial intelligence in video review for the Paris Olympics Taekwondo competition.” Truvace, 2026-11-06. /record/TRV-2026-2398 (accessed on the day you copy it). sha256 0e0e1153b6b0c7dd…
Change history
Every change made to this record since it was first published, and why. Nothing has changed since it went up.
Certified into the record
How to check it yourself, without trusting this page
Download any version from /api/record/TRV-2026-2398 and hash its canonical field yourself, for example with shasum -a 256. You should get its content_hash. Each version’s text also ends with prev:and the hash of the version before it (version 1 points to 64 zeros). If anyone had changed a single character of any version after it was saved, those hashes wouldn’t line up.