This repository contains the implementation of 2 classification models for satellite images:
- K-Nearest Neighbours
- Convolutional Neural Network
The dataset contains 5631 satellite images belonging to 4 classes: cloudy, desert, green area and water.
├── checkpoints
│ └── best_model.pth
├── data
│ ├── cloudy
│ ├── desert
│ ├── green_area
│ └── water
├── README.md
├── run.ipynb
└── src
├── data
│ └── TransformDataset.py
└── utils
├── general_utils.py
└── train_utils.py