Skip to main content

Optical Character Recognition of Cyrillic Characters Using a Convolutional Neural Network

I created a convolutional neural network and a set of complementary utilities for optical character recognition with two classmates in my Introduction to Artificial Intelligence course at Reykjavik University. In particular, I focused on reconstructing water damaged documents written in the Ukrainian language. We used Tensorflow for the neural network and the code is published on GitHub .

Technical Writeup PDF