DriveWorld3D is a framework that integrates temporal video learning, multi‑sensor measurement constraints, and offline generative simulation to build a unified 3D world model for autonomous driving. Its online component tracks road surfaces, volumetric occupancy, dynamic entities, and road relations while explicitly managing physical occupancy, observability, and evidence provenance. The offline component reconstructs scenes with neural radiance fields or 3D Gaussian Splatting and expands rare conditions using generative models, but limits generated content from being used as geometric ground truth.
arXiv:2606. 03159v1 Announce Type: cross Abstract: As autonomous vehicle capabilities advance, the safe evaluation of driving policies in long-tail scenarios remains a critical bottleneck.
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.
SimVTA is a post‑training framework that enhances driving world action models by aligning simulated future videos with candidate trajectories. It samples trajectories, renders high‑reward futures in a real‑world simulator using 3D Gaussian Splatting, and prioritizes video supervision for high‑error futures. The approach improves planning performance, raising PDMS from 86.2 to 91.3 on NAVSIM v1 and achieving strong results on NAVSIM v2 and HUGSIM.
arXiv:2609.22762v1 Announce Type: new Abstract: Generative world-action models (WAMs) jointly generate future video and vehicle actions, while their action branches remain primarily optimized by expe...
arXiv:2603. 28963v2 Announce Type: replace-cross Abstract: Simulation with realistic traffic agents is essential for validating autonomous driving systems.
arXiv:2609.36851v1 Announce Type: new Abstract: End-to-end autonomous driving policies are commonly trained via imitation learning on logged demonstrations without observing the consequences of their...
arXiv:2606. 07366v1 Announce Type: cross Abstract: Self-driving simulations typically rely on data collected in a small number of cities or on hand-authored synthetic scenarios.
arXiv:2606. 16278v1 Announce Type: cross Abstract: Long-tail hazardous scenarios are essential for safety-oriented autonomous driving, yet they are difficult to collect and reproduce at scale.
PanoBench is a new camera‑centric benchmark for panoptic occupancy prediction that unifies 3D geometry, semantics, and instance identities across synthetic and real‑world driving scenes. It combines three datasets—PanoBench‑CarlaOcc, PanoBench‑nuScenes, and PanoBench‑KITTI360—through canonical interfaces and unified evaluation protocols, providing multi‑resolution annotations with persistent instance identities. The benchmark introduces a novel reconstruction pipeline for real‑world data, a unified valid‑mask protocol, and analyses of observation fidelity, geometric completeness, spatial continuity, and temporal consistency, while also benchmarking methods on semantic, panoptic, and 4D occupancy forecasting. whyItMatters":"PanoBench addresses the scarcity of instance‑aware annotations in panoptic occupancy research by offering comprehensive, high‑quality annotations and evaluation protocols that enable more robust development and comparison of autonomous driving perception methods."
Driving world models provide a promising route toward scalable counterfactual data generation and interactive simulation beyond recorded driving logs. Realizing this potential requires a system that c...
arXiv:2511.22187v4 Announce Type: replace Abstract: Realistic and controllable simulation is critical for advancing end-to-end autonomous driving, yet existing approaches often struggle to support no...
HelloWorld is a 2B-parameter driving world model that learns from diverse video data to generate controllable driving scenarios. It uses ego pose, HD maps, and 3D boxes to produce coherent multi‑sensor outputs, including synchronized seven‑camera RGB and conditional LiDAR. The system is designed for efficient repeated inference and is evaluated on visual quality, control fidelity, cross‑view consistency, robustness, and inference speed.