..

LAMP: Localization Aware Multi-camera People Tracking in Metric 3D World

Nan Yang, Julian Straub, Fan Zhang, Richard Newcombe, Jakob Engel, Lingni Ma

LAMP lifts 2D body keypoints from all cameras into a shared 3D world frame using known device pose and calibration, then fits 3D human motion to this ray cloud with a spatio-temporal transformer. This “lift-then-fit” approach sets a new state of the art on both monocular and egocentric benchmarks.