{"ID":2851281,"CreatedAt":"2026-06-01T04:54:23.091178241Z","UpdatedAt":"2026-06-01T04:54:23.091178241Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2510.20726","arxiv_id":"2510.20726","title":"AutoScape: Geometry-Consistent Long-Horizon Scene Generation","abstract":"This paper proposes AutoScape, a long-horizon driving scene generation framework. At its core is a novel RGB-D diffusion model that iteratively generates sparse, geometrically consistent keyframes, serving as reliable anchors for the scene's appearance and geometry. To maintain long-range geometric consistency, the model 1) jointly handles image and depth in a shared latent space, 2) explicitly conditions on the existing scene geometry (i.e., rendered point clouds) from previously generated keyframes, and 3) steers the sampling process with a warp-consistent guidance. Given high-quality RGB-D keyframes, a video diffusion model then interpolates between them to produce dense and coherent video frames. AutoScape generates realistic and geometrically consistent driving videos of over 20 seconds, improving the long-horizon FID and FVD scores over the prior state-of-the-art by 48.6\\% and 43.0\\%, respectively.","short_abstract":"This paper proposes AutoScape, a long-horizon driving scene generation framework. At its core is a novel RGB-D diffusion model that iteratively generates sparse, geometrically consistent keyframes, serving as reliable anchors for the scene's appearance and geometry. To maintain long-range geometric consistency, the mod...","url_abs":"https://arxiv.org/abs/2510.20726","url_pdf":"https://arxiv.org/pdf/2510.20726v1","authors":"[\"Jiacheng Chen\",\"Ziyu Jiang\",\"Mingfu Liang\",\"Bingbing Zhuang\",\"Jong-Chyi Su\",\"Sparsh Garg\",\"Ying Wu\",\"Manmohan Chandraker\"]","published":"2025-10-23T16:44:34Z","proceeding":"cs.CV","tasks":"[\"cs.CV\"]","methods":"[\"Diffusion Model\"]","has_code":false}