Puzzles: Unbounded Video-Depth Augmentation for Scalable End-to-End 3D Reconstruction.

Jiahao Ma, Lei Wang 0108, Miaomiao Liu 0001, David Ahmedt-Aristizabal, Chuong Nguyen