RoboDreamer leverages the power of language to guide video synthesis, enabling robots to plan actions in unseen scenarios. The approach decomposes instructions into manageable parts, promoting generational flexibility and detail in envisioned scenarios…