multi-view-triangulation

This paper introduces Hand Visibility Detector, a dedicated model for per-joint hand visibility estimation, and demonstrates its benefit for multi-view 3D hand pose annotation via visibility-weighted triangulation.