3D Human Pose Estimation with Occlusions: Introducing BlendMimic3D Dataset and GCN Refinement

TL;DR

This paper introduces the BlendMimic3D dataset for occlusion scenarios in 3D human pose estimation and proposes a GCN-based refinement method that improves pose accuracy, especially under occlusions, without retraining existing models.

Contribution

The work presents a new dataset and a GCN refinement module that enhances 3D human pose estimation in occluded conditions, compatible with various existing frameworks.

Findings

GCN refinement improves occluded pose estimation accuracy.

BlendMimic3D dataset effectively mimics real-world occlusion scenarios.

Method achieves comparable results on non-occluded poses.

Abstract

In the field of 3D Human Pose Estimation (HPE), accurately estimating human pose, especially in scenarios with occlusions, is a significant challenge. This work identifies and addresses a gap in the current state of the art in 3D HPE concerning the scarcity of data and strategies for handling occlusions. We introduce our novel BlendMimic3D dataset, designed to mimic real-world situations where occlusions occur for seamless integration in 3D HPE algorithms. Additionally, we propose a 3D pose refinement block, employing a Graph Convolutional Network (GCN) to enhance pose representation through a graph model. This GCN block acts as a plug-and-play solution, adaptable to various 3D HPE frameworks without requiring retraining them. By training the GCN with occluded data from BlendMimic3D, we demonstrate significant improvements in resolving occluded poses, with comparable results for non-occluded ones. Project web page is available at https://blendmimic3d.github.io/BlendMimic3D/.