π⁴ foundation model: one forward pass for 4D scene geometry and motion from single images to 1000-frame videos
rsasaki0109 · x · 2026-11-19
π⁴ is a new foundation model for 4D scene geometry and motion. A single forward pass recovers both geometry and motion, working on inputs ranging from single images to videos with thousands of frames.
More from Research
- ArkenOS launches Optimus-1, billed as the first swarm harness for autonomous research — CatAstro_Piyush · 2026-11-19
- Growing behavior trees as agent harness: an agent that writes its own reflexes — CatAstro_Piyush · 2026-11-19
- MIT's Aran Nayebi: ML conference rebuttals are a language game with no scientific value — aran_nayebi · 2026-11-19
- Stephen Wolfram on the Ruliad: the entangled limit of all possible computation — BecauseCulture · 2026-11-19
- AI bio dev publishes artifact cataloging today's available assays and which CROs offer them — yawnxyz · 2026-11-19
- SEAM benchmark: agents flag harmful actions, then do them anyway 81-89% of the time — Dedap85 · 2026-11-19