Code for the paper - 'Data-Efficient Classification of Radio Galaxies' (https://arxiv.org/abs/2011.13311)
The code is available in the form of jupyter notebooks for training. Here, we provide both versions (original unbalanced dataset, balanced dataset) mentioned in the paper.
Please follow the notebooks to train and test the experiments presented in the paper.
The model_pickle and augmented_model_pickle directory provide individual path in the model.txt file to download the pickle files. Download the models using the given link and place them in individual directory.
In case of testing the model with new samples, use RG_test.py script.