Grounding Image Matching in 3D with MASt3R | VitaVision
PaperECCV 2024 · 2024

Grounding Image Matching in 3D with MASt3R

Vincent LeroyYohann CabonJérôme Revaud
arXiv 2406.09756DOIView in graph
1page built on it
2more pages cite it
3narratives
0registry papers cite it

In the Atlas

ModelPrimary source of
MASt3R

A 3D-grounded image matcher that adds a dense local-descriptor head and an InfoNCE matching loss on top of DUSt3R's pointmap regression, with a fast reciprocal matching scheme, yielding correspondences robust to extreme viewpoint change.

Cited as a reference by 2 pages, grouped by domain

Geometry 2
Feed-Forward 3D ReconstructionVGGT (Visual Geometry Grounded Transformer)

Appears in 3 narratives

NarrativeDepth Becomes General GeometryNarrativeThe Geometry Pipeline CollapsesNarrativeWhere Did the Inductive Bias Go?

Lineage

Papers in the Atlas registry only · one dot per paper
Built upon by · 0
’24
Builds on · 0
Built upon by · 0