TrackEverything: 3D point tracker handles 1000+ frame videos via persistent scene tracks
CSProfKGD · x · 2026-09-27
Ayush Jain and Katerina Fragkiadaki's group introduce TrackEverything, a 3D point tracker that tracks all points across all frames of 1000+ frame videos. The key idea: represent video as persistent 3D scene tracks and initiate new tracks based on new geometry rather than new frames, tying computation to unique 3D scene content instead of redundant 2D pixels for efficient long-horizon dense tracking. Code is available.
Related event: CMU and Meta Release TrackEverything for Long-Video 3D Point Tracking(3 posts)→
More from Research
- Sakana AI × Pathway Post-Transformers lecture on AI and the brain now online — kaixhin · 2026-11-11
- Defending METR: public data can't second-guess its AI speedup findings — peterwildeford · 2026-11-11
- Adversarial RL envs for backdoor insertion vs. detection: where does it converge? — DimitrisPapail · 2026-11-11
- DiffusionEngine: an open-source modular PyTorch DDPM framework seeking contributors — Euphoric-Incident-93 · 2026-11-11
- UniPath (NeurIPS 2026): adaptively coordinating understanding and generation in one model — jindong_wang92 · 2026-11-11
- Versor pitches real2sim2real: robotics sim needs physical dynamics, not just visuals — abhishekunique7 · 2026-11-11