language: en
license: mit
library_name: pytorch
model-index:
- name: baseline
results:
- task:
type: Geoscore
dataset:
name: OSV-5M
type: geolocation
metrics:
- type: geoscore
value: 3361
- task:
type: Haversine Distance
dataset:
name: OSV-5M
type: geolocation
metrics:
- type: haversine distance
value: 1814
- task:
type: Country classification
dataset:
name: OSV-5M
type: geolocation
metrics:
- type: country accuracy
value: 68
- task:
type: Region classification
dataset:
name: OSV-5M
type: geolocation
metrics:
- type: region accuracy
value: 39.4
- task:
type: Area classification
dataset:
name: OSV-5M
type: geolocation
metrics:
- type: area accuracy
value: 10.3
- task:
type: City classification
dataset:
name: OSV-5M
type: geolocation
metrics:
- type: city accuracy
value: 5.9
OpenStreetView-5M
The Many Roads to Global Visual Geolocation 📍🌍
First authors: Guillaume Astruc, Nicolas Dufour, Ioannis Siglidis
Second authors: Constantin Aronssohn, Nacim Bouia, Stephanie Fu, Romain Loiseau, Van Nguyen Nguyen, Charles Raude, Elliot Vincent, Lintao XU, Hongyu Zhou
Last author: Loic Landrieu
Research Institute: Imagine, LIGM, Ecole des Ponts, Univ Gustave Eiffel, CNRS, Marne-la-Vallée, France
Introduction 🌍
OpenStreetView-5M is the first large-scale open geolocation benchmark of streetview images.
To get a sense of the difficulty of the benchmark, you can play our demo.
Our dataset was used in an extensive benchmark of which we provide the best model.
For more details and results, please check out our paper and project page.
Inference 🔥
Our best model on OSV-5M can also be found on huggingface.
First download the repo !git clone https://github.com/gastruc/osv5m.
Then from any script whose cwd is the repos main directory (cd osv5m) run:
from PIL import Image
from models.huggingface import Geolocalizer
geoloc = Geolocalizer.from_pretrained('osv5m/baseline')
img = Image.open('.media/examples/img1.jpeg')
x = geoloc.transform(img).unsqueeze(0) # transform the image using our dedicated transformer
gps = geoloc(x) # B, 2 (lat, lon - tensor in rad)
To reproduce results for this model, run:
python evaluation.py exp=eval_best_model dataset.global_batch_size=1024
Citing 💫
@article{osv5m,
title = {{OpenStreetView-5M}: {T}he Many Roads to Global Visual Geolocation},
author = {Astruc, Guillaume and Dufour, Nicolas and Siglidis, Ioannis
and Aronssohn, Constantin and Bouia, Nacim and Fu, Stephanie and Loiseau, Romain
and Nguyen, Van Nguyen and Raude, Charles and Vincent, Elliot and Xu, Lintao
and Zhou, Hongyu and Landrieu, Loic},
journal = {CVPR},
year = {2024},
}