Search papers, labs, and topics across Lattice.
This paper introduces OmniDreams, a generative world model designed for real-time closed-loop simulation of autonomous vehicles, addressing the limitations of traditional simulators in dynamic environments. By leveraging the Cosmos diffusion model and extensive training on diverse driving scenarios, OmniDreams can generate action-conditioned videos that reflect complex, unobserved phenomena such as extreme weather and unpredictable agent behaviors. Preliminary results demonstrate that a world-action model derived from OmniDreams outperforms existing policy models while using significantly fewer parameters, showcasing its potential as a foundational tool for future autonomous driving systems.
A real-time generative world model can synthesize complex driving scenarios that traditional simulators struggle to capture, enabling safer and more effective evaluation of autonomous vehicle policies.
As autonomous vehicle capabilities advance, the safe evaluation of driving policies in long-tail scenarios remains a critical bottleneck. In closed-loop simulation, the driving policy model actively interacts with the environment, where its actions dynamically update the simulator state and directly influence the next set of generated sensor observations. While recent reconstruction-based neural simulators offer photorealism, they are fundamentally constrained by their initial captured data and struggle to generalize to highly dynamic or novel scenes. To overcome these limitations, we introduce OmniDreams, a foundation generative world model mid- and post-trained from the Cosmos diffusion model to autoregressively generate action-conditioned videos in real time. By leveraging the rich visual priors of Cosmos and mid- and post-training on 21k hours of driving scenarios, OmniDreams synthesizes complex, unobserved phenomena that are hard for traditional simulators to capture, such as extreme weather and unpredictable dynamic agent behaviors. Crucially, it autoregressively conditions its photorealistic sensor generation on past frames, the current simulator state, and immediate driving actions. Deployed in a closed-loop system with the Alpamayo 1 policy model and AlpaSim orchestrator, OmniDreams acts as a highly responsive, reactive environment, providing a scalable and comprehensive solution for training and evaluating next-generation autonomous driving policies. We additionally show preliminary results indicating that a world-action model (WAM) post-trained from OmniDreams achieves strong performance on the Physical AI Autonomous Vehicles NuRec dataset, surpassing the VLA-based Alpamayo 1.5 research policy model while using only 1/5 the total parameters. These results highlight the potential for a real-time world model like OmniDreams to also serve as a backbone for policy architectures.
NVIDIA