Challenge: https://iciar2018-challenge.grand-challenge.org/home/
DataSet: https://drive.google.com/file/d/1B8WqyAHgmcNy9VF_2NF7jrwx4ADUCn3L/view?usp=sharing
Our approach to PART A of the challenge posted above.
We have a custom dataset created "SiameseBaggingDataset" which we use and then we have our tests. The ipynb has been formulated for easy understanding and navigation. NOTE: Since our approach takes a comparison with training data during test(considering a siamese comparison) but WE DO NOT USE TEST LABELS except for generating prediction score.
You can find the project requirements in requirements.txt