Improving esports viewing experience through hierarchical scene detection and tracking

Abstract The role of an observer in esports is to provide spectators with the most engaging scenes in real time. To automate this process, various research has been conducted. In this study, we utilize Vision Transformer (ViT)-based object detection to enhance the accuracy of automatic observers. Ho...

Full description

Saved in:
Bibliographic Details
Main Authors: Ho-Taek Joo, Sung-Ha Lee, Insik Chung, Kyung-Joong Kim
Format: Article
Language:English
Published: Nature Portfolio 2025-03-01
Series:Scientific Reports
Subjects:
Online Access:https://doi.org/10.1038/s41598-025-93692-0
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850057578852122624
author Ho-Taek Joo
Sung-Ha Lee
Insik Chung
Kyung-Joong Kim
author_facet Ho-Taek Joo
Sung-Ha Lee
Insik Chung
Kyung-Joong Kim
author_sort Ho-Taek Joo
collection DOAJ
description Abstract The role of an observer in esports is to provide spectators with the most engaging scenes in real time. To automate this process, various research has been conducted. In this study, we utilize Vision Transformer (ViT)-based object detection to enhance the accuracy of automatic observers. However, while ViT-based detection more accurately identifies engaging game scenes, it often leads to frequent and abrupt scene changes, reducing viewer comfort. To address this issue, we propose a novel hierarchical structure that combines scene detection with scene tracking, maintaining high accuracy while ensuring smoother transitions between scenes. This approach also improves inference speed, as the tracking model is faster than the detection model. We computationally evaluated six observer models in terms of accuracy and camera stability, with our method demonstrating significantly more stable camera control. Additionally, user testing indicated a strong preference for our model over those without tracking. A video comparing our method to the state-of-the-art can be viewed at https://youtu.be/gWiU4GACZEg .
format Article
id doaj-art-7d9d9c73fef3419385f619ee2e776a27
institution DOAJ
issn 2045-2322
language English
publishDate 2025-03-01
publisher Nature Portfolio
record_format Article
series Scientific Reports
spelling doaj-art-7d9d9c73fef3419385f619ee2e776a272025-08-20T02:51:24ZengNature PortfolioScientific Reports2045-23222025-03-0115111310.1038/s41598-025-93692-0Improving esports viewing experience through hierarchical scene detection and trackingHo-Taek Joo0Sung-Ha Lee1Insik Chung2Kyung-Joong Kim3School of Integrated Technology, Gwangju Institute of Science and TechnologyAI Graduate School, Gwangju Institute of Science and TechnologySchool of Integrated Technology, Gwangju Institute of Science and TechnologySchool of Integrated Technology, Gwangju Institute of Science and TechnologyAbstract The role of an observer in esports is to provide spectators with the most engaging scenes in real time. To automate this process, various research has been conducted. In this study, we utilize Vision Transformer (ViT)-based object detection to enhance the accuracy of automatic observers. However, while ViT-based detection more accurately identifies engaging game scenes, it often leads to frequent and abrupt scene changes, reducing viewer comfort. To address this issue, we propose a novel hierarchical structure that combines scene detection with scene tracking, maintaining high accuracy while ensuring smoother transitions between scenes. This approach also improves inference speed, as the tracking model is faster than the detection model. We computationally evaluated six observer models in terms of accuracy and camera stability, with our method demonstrating significantly more stable camera control. Additionally, user testing indicated a strong preference for our model over those without tracking. A video comparing our method to the state-of-the-art can be viewed at https://youtu.be/gWiU4GACZEg .https://doi.org/10.1038/s41598-025-93692-0StarCraftEsportsGame observersSpectators
spellingShingle Ho-Taek Joo
Sung-Ha Lee
Insik Chung
Kyung-Joong Kim
Improving esports viewing experience through hierarchical scene detection and tracking
Scientific Reports
StarCraft
Esports
Game observers
Spectators
title Improving esports viewing experience through hierarchical scene detection and tracking
title_full Improving esports viewing experience through hierarchical scene detection and tracking
title_fullStr Improving esports viewing experience through hierarchical scene detection and tracking
title_full_unstemmed Improving esports viewing experience through hierarchical scene detection and tracking
title_short Improving esports viewing experience through hierarchical scene detection and tracking
title_sort improving esports viewing experience through hierarchical scene detection and tracking
topic StarCraft
Esports
Game observers
Spectators
url https://doi.org/10.1038/s41598-025-93692-0
work_keys_str_mv AT hotaekjoo improvingesportsviewingexperiencethroughhierarchicalscenedetectionandtracking
AT sunghalee improvingesportsviewingexperiencethroughhierarchicalscenedetectionandtracking
AT insikchung improvingesportsviewingexperiencethroughhierarchicalscenedetectionandtracking
AT kyungjoongkim improvingesportsviewingexperiencethroughhierarchicalscenedetectionandtracking