S4VY: Segment Anything in 4D, with consistent identity across space and time
realChrisChoy · x · 2026-10-17
Researchers introduce S4VY (Segment Anything, now in 4D), extending segmentation from 2D to 4D space-time: from any set of RGB observations — a video, scattered frames, or an unordered photo collection — it segments every object with one consistent identity across space and time. Code and checkpoints coming soon.
More from Multimodal
- A 4-minute intro to Invoke's canvas workflow for ComfyUI-averse SD users — no3us · 2026-10-17
- Fable 5.5 Video Covers 40,000 Years of Art History in 15 Seconds — With a Cat Side-Quest — FinanceYF5 · 2026-10-17
- Musician makes hand-drawn 2D music video entirely in code with Claude and Remotion — pizzaisprettyneato · 2026-10-17
- Blind test of 200 real image edits: nano banana pro loses to its cheaper base model — ravisparikh · 2026-10-17
- Striking Midjourney glitch-art cityscapes turn Manhattan into pixel-stretched neon dreams — gizakdag · 2026-10-17
- Creator builds "Claudia x Roadster" film with Claude, MCP, Midjourney and BytePlus — elonmusk · 2026-10-17