---
license: cc-by-4.0
sdk: streamlit
sdk_version: 1.25.0
colorFrom: blue
pinned: false
title: Biomap
emoji: 🐢
colorTo: green
app_file: biomap/streamlit_app.py
---

# Welcome to the project inno-satellite-images-segmentation-gan
![](docs/assets/banner.png)

- **Project name**: inno-satellite-images-segmentation-gan
- **Library name**: library
- **Authors**: Ekimetrics
- **Description**: Segmenting satellite images in a large scale is challenging because grondtruth labels are spurious for medium resolution images (Sentinel 2). We want to improve our algorithm either with data augmentation from a GAN, or to correct or adjust Corine labels. 



## Project Structure
```
- library/ # Your python library
- data/
    - raw/
    - processed/
- docs/
- tests/                            # Where goes each unitary test in your folder
- scripts/                          # Where each automation script will go
- requirements.txt                  # Where you should put the libraries version used in your library
```


## Branch strategy
TBD


## Ethics checklist
TBD



## Starter package
This project has been created using the Ekimetrics Python Starter Package to enforce best coding practices, reusability and industrialization. <br>
If you have any questions please reach out to the inno team and [Théo Alves Da Costa](mailto:theo.alvesdacosta@ekimetrics.com)