MotionForesight: Predicting Future 3D Motion
mangahomanga · x · 2026-07-18
Introducing MotionForesight: repurposing off-the-shelf video models to predict future 3D scene flow after short video context. - Trained on only 40K RGB human interaction videos, yet generalizes to novel home/office scenes. - Core idea: leverage temporal priors from pretrained video models, no language or action labels. - Authors built on TrackCraft3r (a dense 3D tracker on Wan2.1) with lightweight LoRA adaptation, using observed RGB and geometry to predict future object motion.
Related event: MotionForesight predicts future 3D motion from existing video models(6 posts)→
More from Research
- Researcher flags unresolved data transparency issues in compositionality evals — najoungkim · 2026-10-27
- Separate Forms from Estimators: Why Steer by Difference of Means When You Can Use SGD? — johnhewtt · 2026-10-27
- EMNLP paper probes why some LLM biases survive rewording while others don't — radamihalcea · 2026-10-27
- Datology webinar: Curation Studio matched top data baselines with under 20% compute — josh_wills · 2026-10-27
- KANLab: an interactive browser playground for Kolmogorov-Arnold Networks — Lukas Niehaus · 2026-10-27
- CMU's LLM2Cert synthesizes stability certificates verified via SMT and Lean — CarnegieMellonU · 2026-10-27