Homework of Knowledge Discovery Course

The aim of this homework is to understand and build FCN models using PyTorch. The dataset used contains 12,000 synthetically generated images of English digits embedded on random backgrounds (randomly selected from a subset of COCO dataset).
Images are generated with varying fonts, colors, scales and rotations.

This dataset can be found on Kaggle as a benchmark of the work on Effects of Degradations on Deep Neural Network Architectures - GitHub

You can run the code on Google Colab