RoboStereo: Dual-Tower 4D Embodied World Models for Unified Policy Optimization
This addresses the challenge of prohibitive costs and safety risks in real-world interaction for Embodied AI, offering a novel framework for policy optimization, though it appears incremental by building upon existing world model concepts.
The paper tackles the problem of geometric hallucinations and lack of unified optimization in Embodied World Models for scalable Embodied AI, introducing RoboStereo, a dual-tower 4D world model with a unified policy optimization framework that achieves >97% average relative improvement on fine-grained manipulation tasks.
Scalable Embodied AI faces fundamental constraints due to prohibitive costs and safety risks of real-world interaction. While Embodied World Models (EWMs) offer promise through imagined rollouts, existing approaches suffer from geometric hallucinations and lack unified optimization frameworks for practical policy improvement. We introduce RoboStereo, a symmetric dual-tower 4D world model that employs bidirectional cross-modal enhancement to ensure spatiotemporal geometric consistency and alleviate physics hallucinations. Building upon this high-fidelity 4D simulator, we present the first unified framework for world-model-based policy optimization: (1) Test-Time Policy Augmentation (TTPA) for pre-execution verification, (2) Imitative-Evolutionary Policy Learning (IEPL) leveraging visual perceptual rewards to learn from expert demonstrations, and (3) Open-Exploration Policy Learning (OEPL) enabling autonomous skill discovery and self-correction. Comprehensive experiments demonstrate RoboStereo achieves state-of-the-art generation quality, with our unified framework delivering >97% average relative improvement on fine-grained manipulation tasks.