A Modular Multi-stage Lightweight Graph Transformer Network for Human Pose and Shape Estimation from 2D Human Pose

TL;DR

This paper presents a lightweight, modular graph transformer network that efficiently estimates 3D human pose and shape from 2D poses, balancing accuracy with computational speed for practical applications.

Contribution

It introduces a novel multi-stage graph transformer architecture that improves efficiency in human mesh reconstruction from 2D poses, suitable for real-world use.

Findings

Achieves high accuracy with reduced computational complexity

Demonstrates effectiveness on benchmark datasets

Outperforms existing methods in efficiency and accuracy

Abstract

In this research, we address the challenge faced by existing deep learning-based human mesh reconstruction methods in balancing accuracy and computational efficiency. These methods typically prioritize accuracy, resulting in large network sizes and excessive computational complexity, which may hinder their practical application in real-world scenarios, such as virtual reality systems. To address this issue, we introduce a modular multi-stage lightweight graph-based transformer network for human pose and shape estimation from 2D human pose, a pose-based human mesh reconstruction approach that prioritizes computational efficiency without sacrificing reconstruction accuracy. Our method consists of a 2D-to-3D lifter module that utilizes graph transformers to analyze structured and implicit joint correlations in 2D human poses, and a mesh regression module that combines the extracted pose features with a mesh template to produce the final human mesh parameters.