MPL: Lifting 3D Human Pose from Multi-view 2D Poses

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Ghasemzadeh, Seyed Abolfazl, Alahi, Alexandre, De Vleeschouwer, Christophe
Natura: Preprint
Pubblicazione: 2024
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866910571125800960
author Ghasemzadeh, Seyed Abolfazl
Alahi, Alexandre
De Vleeschouwer, Christophe
author_facet Ghasemzadeh, Seyed Abolfazl
Alahi, Alexandre
De Vleeschouwer, Christophe
contents Estimating 3D human poses from 2D images is challenging due to occlusions and projective acquisition. Learning-based approaches have been largely studied to address this challenge, both in single and multi-view setups. These solutions however fail to generalize to real-world cases due to the lack of (multi-view) 'in-the-wild' images paired with 3D poses for training. For this reason, we propose combining 2D pose estimation, for which large and rich training datasets exist, and 2D-to-3D pose lifting, using a transformer-based network that can be trained from synthetic 2D-3D pose pairs. Our experiments demonstrate decreases up to 45% in MPJPE errors compared to the 3D pose obtained by triangulating the 2D poses. The framework's source code is available at https://github.com/aghasemzadeh/OpenMPL .
format Preprint
id arxiv_https___arxiv_org_abs_2408_10805
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle MPL: Lifting 3D Human Pose from Multi-view 2D Poses
Ghasemzadeh, Seyed Abolfazl
Alahi, Alexandre
De Vleeschouwer, Christophe
Computer Vision and Pattern Recognition
Estimating 3D human poses from 2D images is challenging due to occlusions and projective acquisition. Learning-based approaches have been largely studied to address this challenge, both in single and multi-view setups. These solutions however fail to generalize to real-world cases due to the lack of (multi-view) 'in-the-wild' images paired with 3D poses for training. For this reason, we propose combining 2D pose estimation, for which large and rich training datasets exist, and 2D-to-3D pose lifting, using a transformer-based network that can be trained from synthetic 2D-3D pose pairs. Our experiments demonstrate decreases up to 45% in MPJPE errors compared to the 3D pose obtained by triangulating the 2D poses. The framework's source code is available at https://github.com/aghasemzadeh/OpenMPL .
title MPL: Lifting 3D Human Pose from Multi-view 2D Poses
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2408.10805