LingBot Vision Pro
Live 3D scene reconstruction streamed to Apple Vision Pro. Geometry, camera trajectory, and keyframe frustums you walk around in immersive space. Built on LingBot-Map, a Geometric Context Transformer.

Spatial computing has no web demo
Reconstruction and spatial AI work lives in research repos and headset-only demos. There is no way to show the experience to someone without a device, and no simple path from a video feed to a walkable 3D map.
Key Challenges
A live map of the world, in immersive space
We built a visionOS app that streams scene reconstruction from video and renders it as a volumetric, walkable map on Apple Vision Pro. Geometry, camera trajectory, and keyframe frustums grow around you in real time.
Our Approach
Technologies Used
Spatial reconstruction, demo-ready
LingBot Vision Pro turns a research reconstruction pipeline into a presentable visionOS prototype, not a browser product. Public source, plus an immersive walk-around map.
visionOS 2.0+
Runtime
Native Apple Vision Pro prototype (no browser live URL)
2 spaces
Experience modes
Volumetric table-scale map plus full immersive space
Public repo
Source access
GitHub project ready for collaborators and review
Ready for Similar Results?
Tell us about your project goals, blockers, and timeline. We'll scope a clear path from problem to shipped outcome.