Object-Centric World Models from Few-Shot Annotations for Sample-Efficient Reinforcement Learning figure
AlphaXiv 中文论文页面(可滚动查看)