Vincent Lepetit

It is found that, despite performing poorly in zero-shot matching, especially in later layers, feed-forward reconstruction models provide strong representations for linear probing and full matching pipelines, and their raw predictions alone enable competitive matching, albeit only under moderate viewpoint changes and modality gaps.

Gekko turns the challenge of monocular signals into a powerful co-visibility detector, achieving unprecedented accuracy in 3D vision tasks without the need for ground-truth annotations.

FlowObject bridges the gap between generative priors and real-world observations, achieving unprecedented fidelity in 3D reconstruction from sparse views.

Forget static scenes: Paparazzo lets robots actively map *moving* 3D objects with unprecedented accuracy.