CMU and Meta Unveil TrackEverything, a 3D Point Tracker for 1000+ Frame Long Videos
AdamWHarley · x · 2026-09-26
Researchers from Carnegie Mellon University and Meta introduce TrackEverything, a dense 3D point tracker that tracks all points across all frames of long videos (1000+ frames).
- Key idea: tie computation to unique 3D scene content rather than redundant 2D pixels, by de-duplicating 3D scene representations.
- The video is represented as persistent 3D scene tracks in world coordinates, which naturally handles camera motion; per-window static/dynamic classification keeps renderings clean.
- The project page shows extensive dense-tracking visualizations on full-length sequences; the paper is on arXiv and code is coming soon.
- Authors include Ayush Jain, Adam W. Harley, Katerina Fragkiadaki, and Jakob Engel.
Related event: CMU and Meta Release TrackEverything for Long-Video 3D Point Tracking(3 posts)→
More from Multimodal
- Sophon-Realtime-v1, a full-duplex speech-to-speech model, trends on Hugging Face — adventists-ai · 2026-11-11
- Asked Fable 5.2 to 'go wild' making a game, got a Pong clone — and loved it — ariG23498 · 2026-11-11
- Western town built with Fable 5.2: ray-traced weathering ages every board — ravisparikh · 2026-11-11
- Diffusion Week 2026 releases all 10 lectures on video generation, world models and more — LucaAmb · 2026-11-11
- Running Wan 2.2 on a RTX 3080 12GB: 5B Breaks Down, 14B Works Surprisingly Well — fixxxer2606 · 2026-11-11
- Indie dev teases Z-Image-based model 'on par with latest closed-source models' — darlens13 · 2026-11-11