Annotation Techniques for Judo Combat Phase Classification from Tournament Footage

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Miyaguchi, Anthony, Moutahir, Jed, Sutar, Tanmay
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866909422222049280
author Miyaguchi, Anthony
Moutahir, Jed
Sutar, Tanmay
author_facet Miyaguchi, Anthony
Moutahir, Jed
Sutar, Tanmay
contents This paper presents a semi-supervised approach to extracting and analyzing combat phases in judo tournaments using live-streamed footage. The objective is to automate the annotation and summarization of live streamed judo matches. We train models that extract relevant entities and classify combat phases from fixed-perspective judo recordings. We employ semi-supervised methods to address limited labeled data in the domain. We build a model of combat phases via transfer learning from a fine-tuned object detector to classify the presence, activity, and standing state of the match. We evaluate our approach on a dataset of 19 thirty-second judo clips, achieving an F1 score on a $20\%$ test hold-out of 0.66, 0.78, and 0.87 for the three classes, respectively. Our results show initial promise for automating more complex information retrieval tasks using rigorous methods with limited labeled data.
format Preprint
id arxiv_https___arxiv_org_abs_2412_07155
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Annotation Techniques for Judo Combat Phase Classification from Tournament Footage
Miyaguchi, Anthony
Moutahir, Jed
Sutar, Tanmay
Computer Vision and Pattern Recognition
Multimedia
This paper presents a semi-supervised approach to extracting and analyzing combat phases in judo tournaments using live-streamed footage. The objective is to automate the annotation and summarization of live streamed judo matches. We train models that extract relevant entities and classify combat phases from fixed-perspective judo recordings. We employ semi-supervised methods to address limited labeled data in the domain. We build a model of combat phases via transfer learning from a fine-tuned object detector to classify the presence, activity, and standing state of the match. We evaluate our approach on a dataset of 19 thirty-second judo clips, achieving an F1 score on a $20\%$ test hold-out of 0.66, 0.78, and 0.87 for the three classes, respectively. Our results show initial promise for automating more complex information retrieval tasks using rigorous methods with limited labeled data.
title Annotation Techniques for Judo Combat Phase Classification from Tournament Footage
topic Computer Vision and Pattern Recognition
Multimedia
url https://arxiv.org/abs/2412.07155