AI RESEARCH
Real2Edit2Real: Generating Robotic Demonstrations via a 3D Control Interface
arXiv CS.CV
•
ArXi:2512.19402v2 Announce Type: replace-cross Recent progress in robot learning has been driven by large-scale datasets and powerful visuomotor policy architectures, yet policy robustness remains limited by the substantial cost of collecting diverse nstrations, particularly for spatial generalization in manipulation tasks. To reduce repetitive data collection, we present Real2Edit2Real, a framework that generates new nstrations by bridging 3D editability with 2D visual data through a 3D control interface.