AutoSoccerPose: Automated 3D posture Analysis of Soccer Shot Movements

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Yeung, Calvin, Ide, Kenjiro, Fujii, Keisuke
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866929348955602944
author Yeung, Calvin
Ide, Kenjiro
Fujii, Keisuke
author_facet Yeung, Calvin
Ide, Kenjiro
Fujii, Keisuke
contents Image understanding is a foundational task in computer vision, with recent applications emerging in soccer posture analysis. However, existing publicly available datasets lack comprehensive information, notably in the form of posture sequences and 2D pose annotations. Moreover, current analysis models often rely on interpretable linear models (e.g., PCA and regression), limiting their capacity to capture non-linear spatiotemporal relationships in complex and diverse scenarios. To address these gaps, we introduce the 3D Shot Posture (3DSP) dataset in soccer broadcast videos, which represents the most extensive sports image dataset with 2D pose annotations to our knowledge. Additionally, we present the 3DSP-GRAE (Graph Recurrent AutoEncoder) model, a non-linear approach for embedding pose sequences. Furthermore, we propose AutoSoccerPose, a pipeline aimed at semi-automating 2D and 3D pose estimation and posture analysis. While achieving full automation proved challenging, we provide a foundational baseline, extending its utility beyond the scope of annotated data. We validate AutoSoccerPose on SoccerNet and 3DSP datasets, and present posture analysis results based on 3DSP. The dataset, code, and models are available at: https://github.com/calvinyeungck/3D-Shot-Posture-Dataset.
format Preprint
id arxiv_https___arxiv_org_abs_2405_12070
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle AutoSoccerPose: Automated 3D posture Analysis of Soccer Shot Movements
Yeung, Calvin
Ide, Kenjiro
Fujii, Keisuke
Computer Vision and Pattern Recognition
Artificial Intelligence
Image understanding is a foundational task in computer vision, with recent applications emerging in soccer posture analysis. However, existing publicly available datasets lack comprehensive information, notably in the form of posture sequences and 2D pose annotations. Moreover, current analysis models often rely on interpretable linear models (e.g., PCA and regression), limiting their capacity to capture non-linear spatiotemporal relationships in complex and diverse scenarios. To address these gaps, we introduce the 3D Shot Posture (3DSP) dataset in soccer broadcast videos, which represents the most extensive sports image dataset with 2D pose annotations to our knowledge. Additionally, we present the 3DSP-GRAE (Graph Recurrent AutoEncoder) model, a non-linear approach for embedding pose sequences. Furthermore, we propose AutoSoccerPose, a pipeline aimed at semi-automating 2D and 3D pose estimation and posture analysis. While achieving full automation proved challenging, we provide a foundational baseline, extending its utility beyond the scope of annotated data. We validate AutoSoccerPose on SoccerNet and 3DSP datasets, and present posture analysis results based on 3DSP. The dataset, code, and models are available at: https://github.com/calvinyeungck/3D-Shot-Posture-Dataset.
title AutoSoccerPose: Automated 3D posture Analysis of Soccer Shot Movements
topic Computer Vision and Pattern Recognition
Artificial Intelligence
url https://arxiv.org/abs/2405.12070