1 paper · 1 filter
Jai Bardhan, Patrik Drozdik, Josef Sivic +1
Action-conditioned robot world models generate future video frames of the manipulated scene given a robot action sequence, offering a promising alternative for simulating tasks tha…