Mesh-Aware Epipolar Matching for Multi-View Multi-Person 3D Pose Estimation in Basketball

2026-05-28

AI Summary

This work addresses the challenge of multi-person 3D pose estimation in team sports such as basketball, where occlusions, visually similar uniforms, and scarce multi-view annotations hinder performance. The authors propose a training-free cross-view reconstruction framework that, for the first time, integrates dense geometric information from monocular 3D human meshes into epipolar matching. By combining a two-stage matching strategy, union-find clustering, and per-joint triangulation, the method significantly enhances the robustness of cross-view identity association. Evaluated on the SportCenter EPFL and Human-M3 Basketball datasets, the approach achieves MPJPE/PA-MPJPE errors of 59.8/40.7 mm and 74.0/51.8 mm, respectively, outperforming existing training-free methods and demonstrating strong performance under RGB-only input conditions.

Abstract

Multi-view multi-person 3D pose estimation in team sports scenarios remains challenging due to player occlusions, appearance similarity caused by team uniforms, and the scarcity of annotated multi-view data, all of which limit the effectiveness and generalization capability of learning-based methods. In contrast, the performance of training-free approaches is inherently constrained by the accuracy of 2D keypoint detection and the robustness of cross-view association. To address these challenges, we propose Mesh-Aware Epipolar Matching (MAEM), a training-free framework for multi-view multi-person 3D pose estimation. Our method employs a monocular 3D human mesh recovery model as the frontend and introduces a two-stage epipolar matching strategy based on the recovered mesh outputs. Specifically, the proposed framework combines disjoint-set-union-based clustering with per-joint triangulation to achieve robust cross-view association and accurate 3D pose reconstruction. Experiments on two public multi-view basketball datasets demonstrate that MAEM consistently outperforms existing training-free association baselines while achieving competitive RGB-only performance in both indoor and outdoor basketball scenarios. MAEM achieves MPJPE/PA-MPJPE scores of 59.8/40.7 mm on SportCenter EPFL and 74.0/51.8 mm on Human-M3 Basketball, highlighting the effectiveness of dense mesh geometry for cross-view association without requiring target-domain training or fine-tuning.

Problem

• multi-view multi-person 3D pose estimation
• occlusions
• appearance similarity
• cross-view association
• team sports

Innovation

• Mesh-Aware Epipolar Matching
• training-free 3D pose estimation
• multi-view association
• human mesh recovery
• disjoint-set-union clustering