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

Seyed Abolfazl Ghasemzadeh, Alexandre Alahi, Christophe De Vleeschouwer
8/20/2024
cs.CV

Official code: aghasemzadeh/OpenMPL (45 stars, MIT, last commit 2025-12-19).

Abstract

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 .

View on arXivView PDF

Code Implementations(1)

aghasemzadeh/OpenMPLOfficial100%
MIT

Multi-view 3D Pose Lifter

455Shell, PythonJul 15, 202410 months agoMIT

Similar Papers

RUMPL: Ray-Based Transformers for Universal Multi-View 2D to 3D Human Pose Lifting

91% match1 repo222025

DeProPose: Deficiency-Proof 3D Human Pose Estimation via Adaptive Multi-View Fusion

85% match1 repo92025

GST: Precise 3D Human Body from a Single Image with Gaussian Splatting Transformers

84% match1 repo1052024

Towards Robust 3D Pose Transfer with Adversarial Learning

84% match2026

HGMamba: Enhancing 3D Human Pose Estimation with a HyperGCN-Mamba Network

84% match1 repo192025

Towards Robust and Smooth 3D Multi-Person Pose Estimation from Monocular Videos in the Wild

84% match2026

Cite this paper

@article{ghasemzadeh2024mpl,
  title  = {MPL: Lifting 3D Human Pose from Multi-view 2D Poses},
  author = {Seyed Abolfazl Ghasemzadeh and Alexandre Alahi and Christophe De Vleeschouwer},
  year   = {2024},
  eprint = {2408.10805},
  archivePrefix = {arXiv},
  url    = {http://arxiv.org/abs/2408.10805v1}
}

Add a badge to your README

Shows whether official code is linked for this paper.

Code of Paper badge
[![Code of Paper](https://codeofpaper.com/badge/2408.10805.svg)](https://codeofpaper.com/paper/2408.10805)

Discussion