# VOXEL-SAM AR

- **Event:** [Meta Global AI Developer Hackathon](https://cerebralvalley.ai/e/meta-ai-developer-hackathon)
- **When:** Oct 26, 12:00 PM – Nov 4, 9:00 PM (EST)
- **Where:** Online
- **Team:** [Roberto Soto](https://cerebralvalley.ai/u/Naruby)
- **GitHub:** https://github.com/Narubyx/voxel-sam-ar
- **Demo video:** https://youtu.be/YWut0Mld7bM
- **Gallery:** https://cerebralvalley.ai/e/meta-ai-developer-hackathon/hackathon/gallery
- **Page:** https://cerebralvalley.ai/e/meta-ai-developer-hackathon/hackathon/gallery/2716

VOXEL-SAM AR turns a 15-second phone video into a true-scale 3D twin you can place in your room through WebAR — no app install, no LiDAR, no manual modeling.

The problem: creating AR-ready 3D content from real objects today requires expensive scanners or hours of manual work. VOXEL-SAM AR democratizes it: orbit any object with your phone, and the pipeline produces a watertight, correctly-scaled, light-reactive mesh viewable on any modern phone browser.

Pipeline: (1) 360° video upload → GPU segmentation isolates the object across frames; (2) a vision-language model infers true 3D geometry, color, and scale from the frames; (3) an agentic loop iteratively repairs the mesh until it passes watertight/manifold checks; (4) the result ships as GLB+USDZ to a WebAR viewer with QR.

Demo result: a red ceramic mug reconstructed at its exact real height (76.2mm), with hollow interior, C-handle hole, and glossy PBR material that reacts to light — side-by-side with the real object in AR.

## More from Meta Global AI Developer Hackathon

- [StudyHub](https://cerebralvalley.ai/e/meta-ai-developer-hackathon/hackathon/gallery/2713)
- [4Dimension](https://cerebralvalley.ai/e/meta-ai-developer-hackathon/hackathon/gallery/2714)
- [Lift Africa Tech](https://cerebralvalley.ai/e/meta-ai-developer-hackathon/hackathon/gallery/2715)
- [BioDigit](https://cerebralvalley.ai/e/meta-ai-developer-hackathon/hackathon/gallery/2717)
- [Layer_Zero](https://cerebralvalley.ai/e/meta-ai-developer-hackathon/hackathon/gallery/2718)
- [Barghest](https://cerebralvalley.ai/e/meta-ai-developer-hackathon/hackathon/gallery/2719)

---

Markdown version of https://cerebralvalley.ai/e/meta-ai-developer-hackathon/hackathon/gallery/2716. Site index for agents: https://cerebralvalley.ai/llms.txt · full text: https://cerebralvalley.ai/llms-full.txt