Multi Label Segmentation using U-NET on Clouds

Steps to execute :-

1. Download the input dataset stored Here
2. Unzip the 'input' folder.
3. Download the models from Here
4. Place the models in 'input/train_images'.

References :-

• https://medium.com/@juliawu/fast-ai-lesson-3-b462367fe1ec
• https://towardsdatascience.com/fastai-image-segmentation-eacad8543f6f
• https://forums.fast.ai/t/unet-how-to-get-4-channel-output/54674/4
• https://matplotlib.org/3.1.1/gallery/text_labels_and_annotations/custom_legends.html
• https://forums.fast.ai/t/how-to-load-multiple-classes-of-rle-strings-from-csv-severstal-steel-competition/51445