IPEO-Understanding-the-Amazon-from-Space

This project is part of the ENV-540 Image Processing for Earth Observation course at EPFL and aims to predict multiple labels on satellite images of the Amazon river and its surroundings. This is done by training a convolutional neural network on a dataset of 40479 labeled images.

Authors

Louis Poulain--Auzéau, [email protected]

Basile Tornare, [email protected]

Octavio Profeta, [email protected]

Data

Download the IPEO_Planet_project folder don't already have it from here:
https://drive.google.com/drive/folders/1tOMxGHMRtY8E1p1NKun6Wi_4DHMmRjAq?usp=sharing

Folder setup

The structure should be as follows:

• Submission_folder
  • IPEO-Understanding-the-Amazon-from-Space
    • some code + logs + evaluation.ipynb
  • IPEO_Planet_project
    • checkpoints
    • train_labels.csv
    • train-jpg

Running

Create a local virtual environment in the venv folder

python -m venv venv

Activate the newly created environment

source venv/bin/activate

Install requirements

pip install -r requirements.txt

Here you need to download the file containing data (cf above for the link)

cd in the good directory

cd IPEO-Understanding-the-Amazon-from-Space

Start a jupyter session

jupyter notebook  

Run evaluation.ipynb

When you are done, close the venv

deactivate