Fetching the paper…

SNAP: Self-Supervised Neural Maps for Visual Positioning and Semantic Understanding · Around