I've been wanting to make 3D reconstructions not just realistic, but also **interactable** and **actionable** for years. Thanks to @XHongchi97338, we're now a step closer! Introducing DRAWER — a framework for the automatic construction of realistic, interactive digital twins.

DRAWER integrates the strengths of GS, Neural SDF, physical reasoning from foundation models, and generative models, enabling the system to build a comprehensive and nuanced understanding of the scene. Some nice features of DRAWER: ✅Geometry ✅Appearance ✅Articulation ✅Speed

No need to interact with the environment during the capture! DRAWER works with just a static video input.

Check out @XHongchi97338's post or our project page for more details! Webpage: Joint w/ researchers from UIUC, @uwcse, @allen_ai, @CornellCIS @EntongSu @memmelma @prodarhan @yu_raymond5 Numfor Ali Farhadi (@RAIVNLab) @abhishekunique7 @ShenlongWang

@uwcse @allen_ai @CornellCIS @EntongSu @memmelma @prodarhan @yu_raymond5 @RAIVNLab @abhishekunique7 @ShenlongWang Some cool applications enabled by DRAWER: ✅Interactive shooting games ✅Real-to-Sim-to-Real

@XHongchi97338 Wait a second, I recognize that kitchen!