PLONK_YFCC_flow

This is a model of the approach described in the paper "Around the World in 80 Timesteps: A Generative Approach to Global Visual Geolocation"

@misc{dufour2024world80timestepsgenerative,
      title={Around the World in 80 Timesteps: A Generative Approach to Global Visual Geolocation}, 
      author={Nicolas Dufour and David Picard and Vicky Kalogeiton and Loic Landrieu},
      year={2024},
      eprint={2412.06781},
      archivePrefix={arXiv},
      primaryClass={cs.CV},
      url={https://arxiv.org/abs/2412.06781}, 
}

• Code: https://github.com/nicolas-dufour/plonk
• Docs: The model can be simply run by doing:

pip install diff-plonk

from plonk import PLONKPipeline

pipeline = PLONKPipeline.from_pretrained("nicolas-dufour/PLONK_YFCC_flow")

gps_coords = pipeline(images, batch_size=1024)