AI RESEARCH
MASt3R-Nav: WayPixel Navigation in Relative 3D Maps
arXiv CS.AI
•
ArXi:2605.24111v1 Announce Type: cross Visual navigation ability is strongly tied to its underlying representation of the world. Unlike classical 3D maps that require globally-consistent geometry, image- or object-relative topological graphs almost entirely do away with geometric understanding. But, this comes at the cost of navigation capability, often limiting it to merely teach-and-repeat. In this work, we propose a novel map representation in the form of pixel-relative connectivity, which is geometrically accurate but does not require global geometric consistency.