This contains the code for the project - Creating an Interpretable Classifier.

1. Train the encoder-decoder network using train_semseg.ipynb
2. Create the decision tree model from decision_tree.ipynb
3. Evaluate the complete pipeline using evaluate.ipynb
4. Compare the above computed accuracy and F1 score from other standard classifiers: vgg16.ipynb, alexnet.ipynb

1. Pretrained ENet model: link
2. Pretrained VGG16 classifier: link
3. Pretrained AlexNet Classifier: link
4. Decision tree classifier: link