TensorX
返回文献探索

Paper · arXiv 2503.18945

Aether: Geometric-Aware Unified World Modeling

Aether Team, Haoyi Zhu, Yifan Wang, Jianjun Zhou, Wenzheng Chang, Yang Zhou, Zizun Li, Junyi Chen, Chunhua Shen, Jiangmiao Pang, Tong He

28 upvotesMarch 24, 2025arXiv 预印本
AI 摘要

Aether is a unified framework that combines 4D dynamic reconstruction, action-conditioned video prediction, and goal-conditioned visual planning, achieving zero-shot generalization and superior reconstruction without real-world data.

4D dynamic reconstructionaction-conditioned video predictiongoal-conditioned visual planningtask-interleaved feature learningsynthetic-to-real generalizationzero-shot generalizationintrinsic geometric modelinggeometry-informed action spaceautonomous trajectory planning

Abstract

The integration of geometric reconstruction and generative modeling remains a critical challenge in developing AI systems capable of human-like spatial reasoning. This paper proposes Aether, a unified framework that enables geometry-aware reasoning in world models by jointly optimizing three core capabilities: (1) 4D dynamic reconstruction, (2) action-conditioned video prediction, and (3) goal-conditioned visual planning. Through task-interleaved feature learning, Aether achieves synergistic knowledge sharing across reconstruction, prediction, and planning objectives. Building upon video generation models, our framework demonstrates unprecedented synthetic-to-real generalization despite never observing real-world data during training. Furthermore, our approach achieves zero-shot generalization in both action following and reconstruction tasks, thanks to its intrinsic geometric modeling. Remarkably, even without real-world data, its reconstruction performance far exceeds that of domain-specific models. Additionally, Aether leverages a geometry-informed action space to seamlessly translate predictions into actions, enabling effective autonomous trajectory planning. We hope our work inspires the community to explore new frontiers in physically-reasonable world modeling and its applications.

北京市昌平区探索星信息技术及软件开发工作室

京ICP备2026059466号