Update README.md

README.md

@@ -55,38 +55,27 @@ model-index:
       value: 5.9
 ---
 
-# Geolocation baseline on OSV-5M
-More details to be released upon publication (\<tbr\>).
-Everything is based on the OSV-5M benchmark dataset.
+![image/png](https://cdn-uploads.huggingface.co/production/uploads/654bb2591a9e65ef2598d8c4/0Z-GMa6SSLgXFmrplC0WD.png)
 
-## Model Details
+# OpenStreetView-5M <br><sub>The Many Roads to Global Visual Geolocation 📍🌍</sub>
 
-\<tbr\>
+**First authors:** [Guillaume Astruc](https://gastruc.github.io/), [Nicolas Dufour](https://nicolas-dufour.github.io/), [Ioannis Siglidis](https://imagine.enpc.fr/~siglidii/)  
+**Second authors:** [Constantin Aronssohn](), Nacim Bouia, [Stephanie Fu](https://stephanie-fu.github.io/), [Romain Loiseau](https://romainloiseau.fr/), [Van Nguyen Nguyen](https://nv-nguyen.github.io/), [Charles Raude](https://imagine.enpc.fr/~raudec/), [Elliot Vincent](https://imagine.enpc.fr/~vincente/), Lintao XU, Hongyu Zhou  
+**Last author:** [Loic Landrieu](https://loiclandrieu.com/)  
+**Research Institute:** [Imagine](https://imagine.enpc.fr/), _LIGM, Ecole des Ponts, Univ Gustave Eiffel, CNRS, Marne-la-Vallée, France_  
 
-### Model Description
+## Introduction 🌍
+[OpenStreetView-5M](https://huggingface.co/datasets/osv5m/osv5m) 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](https://huggingface.co/spaces/osv5m/plonk).  
+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](arxiv) and [project page](https://imagine.enpc.fr/~guillaume-astruc/osv-5m).  
 
-\<tbr\>
+### Inference 🔥
+![image/png](https://cdn-uploads.huggingface.co/production/uploads/654bb2591a9e65ef2598d8c4/mmTZy5ELTwLiLap8pO4xV.png)
 
-- **Developed by:** \<tbr\>
-- **License:** mit
-- **Based on hf models:** \<tbr\>
-
-### Model Sources [optional]
-
-<!-- Provide the basic links for the model. -->
-
-- **Repository:** \<tbr\>
-- **Paper:** \<tbr\>
-- **Human Evaluation** \<tbr\>
-
-## Usage
-
-The main purpose of this model is academic usage. We provide a hugging face repo both to facilitate accessing and run inference to our model.
-
-### Example usage
-
-First download the repo `!git clone <tbr>`.
-Then from any script whose `cwd` is the repos main directory (`cd <tbr>`) run:
+Our best model on OSV-5M can also be found on [huggingface](https://huggingface.co/osv5m/baseline).  
+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:
 
 ```python
 from PIL import Image
@@ -96,4 +85,24 @@ 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:
+
+```bash
+python evaluation.py exp=eval_best_model dataset.global_batch_size=1024
+```
+
+### Citing 💫
+
+```bibtex
+@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},
+  }
 ```