arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2603.04553cs.LG
收藏

潜在粒子世界模型:自监督的以对象为中心的随机动态建模

Latent Particle World Models: Self-supervised Object-centric Stochastic Dynamics Modeling

Carnegie Mellon University(卡内基梅隆大学);UT Austin(德克萨斯大学奥斯汀分校);Brown University(布朗大学);Lambda;Technion(技术学院)

Tal Daniel, Carl Qi, Dan Haramati, Amir Zadeh, Chuan Li, Aviv Tamar, Deepak Pathak, David Held

AI总结:

LPWM通过自监督学习实现现实世界多对象数据的随机动态建模,并在决策制定中展现卓越性能。

摘要:

我们介绍了潜在粒子世界模型(LPWM),一种自监督的以对象为中心的世界模型,能够扩展到现实世界多对象数据集,并适用于决策制定。LPWM能够自主发现关键点、边界框和对象掩码,直接从视频数据中学习丰富的场景分解,而无需监督。我们的架构是端到端地从视频中训练的,支持灵活的行动、语言和图像目标条件。LPWM通过新颖的潜在动作模块建模随机粒子动力学,并在多样化的现实世界和合成数据集上取得了最先进的结果。除了随机视频建模之外,LPWM还能够迅速应用于决策制定,包括目标条件的模仿学习,如我们在论文中所展示的。代码、数据、预训练模型和视频回放可供参考:https://taldatech.github.io/lpwm-web

英文摘要:

We introduce Latent Particle World Model (LPWM), a self-supervised object-centric world model scaled to real-world multi-object datasets and applicable in decision-making. LPWM autonomously discovers keypoints, bounding boxes, and object masks directly from video data, enabling it to learn rich scene decompositions without supervision. Our architecture is trained end-to-end purely from videos and supports flexible conditioning on actions, language, and image goals. LPWM models stochastic particle dynamics via a novel latent action module and achieves state-of-the-art results on diverse real-world and synthetic datasets. Beyond stochastic video modeling, LPWM is readily applicable to decision-making, including goal-conditioned imitation learning, as we demonstrate in the paper. Code, data, pre-trained models and video rollouts are available: https://taldatech.github.io/lpwm-web

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑