3D scanning might be over. 32 photos - is all it took to turn NVIDIA's headquarters into…
This is a AI post classified by Jev as AI research (a model release), kept by the AI Radar because it carries real work, not commentary.
3D scanning might be over. 32 photos - is all it took to turn NVIDIA's headquarters into a real-time flyable 3D scene. Their new Atlas model is one network trained natively on text, images, video and 3D. It was trained on NVIDIA Blackwell GPUs, so it is flying through the halls of its own hardware. Photos go in as spatial context, the camera path goes in as geometry, and it generates the missing views with pixel-perfect control. On sparse-view reconstruction they report 25.3 pointmap error against 28.7 for the best specialist model. It will power the next Marble versions. For now there is no
Posted by Stefan 3D AI (8.5k followers) 1 days ago · 134 likes · 10.7k views · view the original post on X. Kept by the AI Radar as AI research. Tools mentioned: worldlabs.ai.
More AI work like this
- SoupFold from KAIST: no single co-folding model wins everywhere, so learn simple… — @biogerontology
- We built AI that actually knows aging biology. Longevity-LLM (9B params) beats OpenAI… — @biogerontology
- We need more examples like this in the open-source RL ecosystem — @adithya_s_k
- "Verbalizing Subliminal Learning Effects Using Text Optimization" — @askalphaxiv
- “What Does Privileged Information Add to On-Policy Self-Distillation?” — @askalphaxiv
- “Score Centering Stabilizes Off-Policy Reinforcement Learning” — @askalphaxiv
- Banger paper from Google Cloud AI Research. — @dair_ai
- If you liked the MiMo-V2.6 livestreamed RL run, don’t forget to check out Marin’s… — @adi_baradwaj
Every post is read and classified by Jev (TypeSafe): what it is, which market it belongs to, and whether the link is a real tool. 27.2k posts from 4.8k X accounts over the last 14 days, 1.2k tools, 19 markets. Collected every 5 minutes, fully re-ranked every hour — last update 2026-09-19 18:45 UTC. Full method.