Grounding World Simulation Models in a Real-World Metropolis
This work addresses the challenge of creating realistic, dynamic simulations of actual urban environments for applications in urban planning, autonomous systems, or virtual reality, representing a novel method for a known bottleneck rather than a foundational breakthrough.
The authors tackled the problem of generating city-scale world models grounded in real cities by developing Seoul World Model (SWM), which uses retrieval-augmented conditioning and novel techniques like cross-temporal pairing and a Virtual Lookahead Sink to outperform existing methods in generating spatially faithful, temporally consistent, long-horizon videos over trajectories of hundreds of meters.
What if a world simulation model could render not an imagined environment but a city that actually exists? Prior generative world models synthesize visually plausible yet artificial environments by imagining all content. We present Seoul World Model (SWM), a city-scale world model grounded in the real city of Seoul. SWM anchors autoregressive video generation through retrieval-augmented conditioning on nearby street-view images. However, this design introduces several challenges, including temporal misalignment between retrieved references and the dynamic target scene, limited trajectory diversity and data sparsity from vehicle-mounted captures at sparse intervals. We address these challenges through cross-temporal pairing, a large-scale synthetic dataset enabling diverse camera trajectories, and a view interpolation pipeline that synthesizes coherent training videos from sparse street-view images. We further introduce a Virtual Lookahead Sink to stabilize long-horizon generation by continuously re-grounding each chunk to a retrieved image at a future location. We evaluate SWM against recent video world models across three cities: Seoul, Busan, and Ann Arbor. SWM outperforms existing methods in generating spatially faithful, temporally consistent, long-horizon videos grounded in actual urban environments over trajectories reaching hundreds of meters, while supporting diverse camera movements and text-prompted scenario variations.