Jason AI
Mini toolsD4RT
By DeepMind
New
Image
Gen 4
Overview
D4RT is DeepMind’s unified 4D scene reconstruction and tracking model that turns ordinary videos into a fast, queryable representation of 3D geometry and motion, solving tracking, depth and pose up to hundreds of times faster than prior work.
Description
D4RT (Dynamic 4D Reconstruction and Tracking) is DeepMind’s unified encoder-decoder Transformer that converts video into a queryable 4D representation of geometry and motion. It can answer where any pixel lies in 3D space over time and camera views, so tasks like point tracking, depth, camera pose and point-cloud reconstruction become simple queries, achieving state-of-the-art accuracy 18-300× faster than prior systems.
About DeepMind
DeepMind is a technology company that specializes in artificial intelligence and machine learning.
View Company Profile
Last updated: January 27, 2026
