Zero-manual-annotation multi-view 3D label-production pipeline for high-occlusion contact sports, plus the monocular lifting recipe built on it — with a layer-wise error budget of what each stage costs
computer-vision weak-supervision human-pose-estimation occlusion 3d-human-pose multi-view combat-sports pseudo-labels motionbert monocular-3d harmony4d label-pipeline
-
Updated
Oct 6, 2026 - Python