Sat2Scene: 3D Urban Scene Generation from Satellite Images with Diffusion
This addresses the challenge of creating adaptable 3D scenes for applications like games and maps, though it is incremental by building on existing diffusion and neural rendering methods.
The paper tackles the problem of generating 3D urban scenes from satellite images by proposing a novel architecture that uses diffusion models with 3D sparse representations and neural rendering, resulting in photo-realistic street-view sequences and cross-view scenes as demonstrated on city-scale datasets.
Directly generating scenes from satellite imagery offers exciting possibilities for integration into applications like games and map services. However, challenges arise from significant view changes and scene scale. Previous efforts mainly focused on image or video generation, lacking exploration into the adaptability of scene generation for arbitrary views. Existing 3D generation works either operate at the object level or are difficult to utilize the geometry obtained from satellite imagery. To overcome these limitations, we propose a novel architecture for direct 3D scene generation by introducing diffusion models into 3D sparse representations and combining them with neural rendering techniques. Specifically, our approach generates texture colors at the point level for a given geometry using a 3D diffusion model first, which is then transformed into a scene representation in a feed-forward manner. The representation can be utilized to render arbitrary views which would excel in both single-frame quality and inter-frame consistency. Experiments in two city-scale datasets show that our model demonstrates proficiency in generating photo-realistic street-view image sequences and cross-view urban scenes from satellite imagery.