Multi-View Person Matching and 3D Pose Estimation with Arbitrary Uncalibrated Camera Networks

TL;DR

This paper introduces PME, a novel method for cross-view person matching and 3D pose estimation that does not require camera calibration or 3D training data, using clustering, triangulation, and bundle adjustment.

Contribution

Proposes PME, a calibration-free approach for multi-view person matching and 3D pose estimation, leveraging clustering and geometric constraints without needing 3D data or camera poses.

Findings

Outperforms existing methods in cross-view person matching

Achieves state-of-the-art 3D pose estimation without calibration or 3D training data

Demonstrates strong generalization across diverse datasets

Abstract

Cross-view person matching and 3D human pose estimation in multi-camera networks are particularly difficult when the cameras are extrinsically uncalibrated. Existing efforts generally require large amounts of 3D data for training neural networks or known camera poses for geometric constraints to solve the problem. However, camera poses and 3D data annotation are usually expensive and not always available. We present a method, PME, that solves the two tasks without requiring either information. Our idea is to address cross-view person matching as a clustering problem using each person as a cluster center, then obtain correspondences from person matches, and estimate 3D human poses through multi-view triangulation and bundle adjustment. We solve the clustering problem by introducing a "size constraint" using the number of cameras and a "source constraint" using the fact that two people from the same camera view should not match, to narrow the solution space to a small feasible region. The 2D human poses used in clustering are obtained through a pre-trained 2D pose detector, so our method does not require expensive 3D training data for each new scene. We extensively evaluate our method on three open datasets and two indoor and outdoor datasets collected using arbitrarily set cameras. Our method outperforms other methods by a large margin on cross-view person matching, reaches SOTA performance on 3D human pose estimation without using either camera poses or 3D training data, and shows good generalization ability across five datasets of various environment settings.