GDAlab/GeoContext-v1

Geospecific View Generation - Geometry-Context Aware High-resolution Ground View Inference from Satellite Views

🌐 Homepage | 📖 arXiv

Introduction

Predicting realistic ground views from satellite imagery in urban scenes is a challenging task due to the significant view gaps between satellite and ground-view images. We propose a novel pipeline to tackle this challenge, by generating geospecifc views that maximally respect the weak geometry and texture from multi-view satellite images. Different from existing approaches that hallucinate images from cues such as partial semantics or geometry from overhead satellite images, our method directly predicts ground-view images at geolocation by using a comprehensive set of information from the satellite image, resulting in ground-level images with a resolution boost at a factor of ten or more. We leverage a novel building refinement method to reduce geometric distortions in satellite data at ground level, which ensures the creation of accurate conditions for view synthesis using diffusion networks. Moreover, we proposed a novel geospecific prior, which prompts distribution learning of diffusion models to respect image samples that are closer to the geolocation of the predicted images. We demonstrate our pipeline is the first to generate close-to-real and geospecific ground views merely based on satellite images.

Description

The GeoContext-v1 contains 4463 pairs of satellite-ground data. Including:

• satellite_rgb: [512x512x3] satellite rgb texture, generated by orthophoto projection from satellite textured mesh.
• satellite_semantics: [512,512,3] satellite semantics containing two class, where "ground" as [120,120,70], "building" as [180,120,120]
• ground_rgb: [256x512x3] ground-view satellite rgb texture, generated by panoramic projection from satellite textured mesh in Blender.
• ground_semantics: [256x512x3] ground-view satellite semantics containing three class, where "ground" "building" same as "satellite_semantics" and "sky" as [6,230,230]
• ground_depth: [256x512x1] ground-view satellite depth, generated in same way as "ground_rgb"
• gt: [256x512x3] ground truth ground-view rgb, downloaded from google street 360.

Citation

BibTex:

@misc{xu2024geospecificviewgeneration,
      title={Geospecific View Generation -- Geometry-Context Aware High-resolution Ground View Inference from Satellite Views}, 
      author={Ningli Xu and Rongjun Qin},
      year={2024},
      eprint={2407.08061},
      archivePrefix={arXiv},
      primaryClass={cs.CV},
      url={https://arxiv.org/abs/2407.08061}, 
}