Contents

Important Note

• Start Date: September 2022
• This is a work in progress. This is a code first learning by doing method. This page will be updated as the project progresses and realted learnings will be shared in the blog section of the website. The Appendix section contains the blog section links, other citings for the learnings & referal material.
• End Date: NA

GitHub Link for the Project

• The code, data & any additional utilities will be hosted on my github account and will be updated as worked on.
• [GitHub Link] (https://github.com/RaviChandraVeeramachaneni/recognizing-handwritten-academic-documents)

Goal

Problem statement

Phase 1

• The goal of the project is to recognize the handwritten academic documents that contain the text along with the complex algorithmic equations.
• Sample Image looks like below

Phase 2

• After recognizing the text, if there are contextual errors for example a mistake in spelling. Based on the Natural language understanding / processing, correct the spelling and present to the user.

Tech Stack Used

• OCR (Optical Character Recognition)

Phase 1

Data Gathering

• The initial data that is used for the project is gathered from Dr. Arash Rafiey, Math & Computer Science department, Indiana State University.
• The dataset will be added continuously.
• The dataset is a collection of handwritten academic documents and is hosted on [github] (https://github.com/RaviChandraVeeramachaneni/recognizing-handwritten-academic-documents/tree/main/dataset)

Data Cleaning

Appendix