1 paper · 1 filter
Mingju Gao, Qingle Liu, Yuzhao Peng +8
Video world models can produce visually convincing yet physically inconsistent sequences, raising concerns about their reliability for prediction and planning in embodied AI system…