GPS as a Control Signal for Image Generation

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Feng, Chao, Chen, Ziyang, Holynski, Aleksander, Efros, Alexei A., Owens, Andrew
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866913660869279744
author Feng, Chao
Chen, Ziyang
Holynski, Aleksander
Efros, Alexei A.
Owens, Andrew
author_facet Feng, Chao
Chen, Ziyang
Holynski, Aleksander
Efros, Alexei A.
Owens, Andrew
contents We show that the GPS tags contained in photo metadata provide a useful control signal for image generation. We train GPS-to-image models and use them for tasks that require a fine-grained understanding of how images vary within a city. In particular, we train a diffusion model to generate images conditioned on both GPS and text. The learned model generates images that capture the distinctive appearance of different neighborhoods, parks, and landmarks. We also extract 3D models from 2D GPS-to-image models through score distillation sampling, using GPS conditioning to constrain the appearance of the reconstruction from each viewpoint. Our evaluations suggest that our GPS-conditioned models successfully learn to generate images that vary based on location, and that GPS conditioning improves estimated 3D structure.
format Preprint
id arxiv_https___arxiv_org_abs_2501_12390
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle GPS as a Control Signal for Image Generation
Feng, Chao
Chen, Ziyang
Holynski, Aleksander
Efros, Alexei A.
Owens, Andrew
Computer Vision and Pattern Recognition
We show that the GPS tags contained in photo metadata provide a useful control signal for image generation. We train GPS-to-image models and use them for tasks that require a fine-grained understanding of how images vary within a city. In particular, we train a diffusion model to generate images conditioned on both GPS and text. The learned model generates images that capture the distinctive appearance of different neighborhoods, parks, and landmarks. We also extract 3D models from 2D GPS-to-image models through score distillation sampling, using GPS conditioning to constrain the appearance of the reconstruction from each viewpoint. Our evaluations suggest that our GPS-conditioned models successfully learn to generate images that vary based on location, and that GPS conditioning improves estimated 3D structure.
title GPS as a Control Signal for Image Generation
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2501.12390