Multi-Person 3D Pose and Shape Estimation via Inverse Kinematics and Refinement

TL;DR

This paper introduces a novel coarse-to-fine pipeline for multi-person 3D mesh estimation from monocular images, combining robust skeleton estimation, inverse kinematics, and Transformer-based refinement to handle occlusions and interactions.

Contribution

It presents a new pipeline integrating occlusion-robust skeleton estimation, inverse kinematics, and relation-aware Transformer refinement for multi-person 3D mesh reconstruction.

Findings

Outperforms state-of-the-art on 3DPW, MuPoTS, and AGORA datasets.

Effectively handles occlusions and person interactions.

Demonstrates significant accuracy improvements over existing methods.

Abstract

Estimating 3D poses and shapes in the form of meshes from monocular RGB images is challenging. Obviously, it is more difficult than estimating 3D poses only in the form of skeletons or heatmaps. When interacting persons are involved, the 3D mesh reconstruction becomes more challenging due to the ambiguity introduced by person-to-person occlusions. To tackle the challenges, we propose a coarse-to-fine pipeline that benefits from 1) inverse kinematics from the occlusion-robust 3D skeleton estimation and 2) Transformer-based relation-aware refinement techniques. In our pipeline, we first obtain occlusion-robust 3D skeletons for multiple persons from an RGB image. Then, we apply inverse kinematics to convert the estimated skeletons to deformable 3D mesh parameters. Finally, we apply the Transformer-based mesh refinement that refines the obtained mesh parameters considering intra- and inter-person relations of 3D meshes. Via extensive experiments, we demonstrate the effectiveness of our method, outperforming state-of-the-arts on 3DPW, MuPoTS and AGORA datasets.