PoseGU: 3D Human Pose Estimation with Novel Human Pose Generator and Unbiased Learning

TL;DR

PoseGU introduces a novel human pose generator combined with unbiased learning to enhance 3D pose estimation, especially in generalizing to unseen poses with limited training data, outperforming existing methods.

Contribution

The paper presents PoseGU, a new pose generator and unbiased learning framework that improves diversity and generalization in 3D human pose estimation from small datasets.

Findings

Outperforms state-of-the-art methods on benchmark datasets

Generates more diverse 3D poses

Achieves better generalization to unseen poses

Abstract

3D pose estimation has recently gained substantial interests in computer vision domain. Existing 3D pose estimation methods have a strong reliance on large size well-annotated 3D pose datasets, and they suffer poor model generalization on unseen poses due to limited diversity of 3D poses in training sets. In this work, we propose PoseGU, a novel human pose generator that generates diverse poses with access only to a small size of seed samples, while equipping the Counterfactual Risk Minimization to pursue an unbiased evaluation objective. Extensive experiments demonstrate PoseGU outforms almost all the state-of-the-art 3D human pose methods under consideration over three popular benchmark datasets. Empirical analysis also proves PoseGU generates 3D poses with improved data diversity and better generalization ability.