PhiZero: A World Model Built Around Physical Language
A physical world model that learns a compact discrete 'physical language' from in-the-wild videos, then uses a reason-then-render paradigm to generate and understand physically coherent video
2 posts tagged with "computer-vision"
A physical world model that learns a compact discrete 'physical language' from in-the-wild videos, then uses a reason-then-render paradigm to generate and understand physically coherent video
通过旋转特征实现连续分布式对象中心表示，将对象发现从简单玩具数据扩展到真实世界数据