The project is was inspired from the original U-Net Architecture paper "U-Net: Convolutional Networks for Biomedical Image Segmentation"[https://arxiv.org/abs/1505.04597]
- The Architecture of the model implemented can be found in the following Image
Downloading from Kaggle
kaggle competitions download -c carvana-image-masking-challenge
- Extract the dataset
- create a New folder "valid"
- select copy and paste the last 48 Images in the train folder to "valid"
Clone the project
git clone darthvader2/U-Net-implementation-from-scratch
Go to the project directory
cd U-Net-implementation-from-scratch
Install dependencies
pip install -r requirements.txt
Training the model
python train.py