Project Idea | noteSort (Classify handwritten notes)

Project Name - noteSort Project Description - An app that will classify the images of handwritten class notes according to different subjects so that they are convenient to manage and read. Introduction One real problem for the people who go by their own ambitions in institutions is having to plead, or possibly threaten, others to give them class notes. Even after getting the notes, having to manage, organize, and making sense of them a day before the exams is quite a pain. Using OCR and machine learning, noteSort will allow one get a "table of contents" for the notes so that it becomes like a book and is convenient to go through. Suggested (online) reading material will also be referenced based on the classification. The user can choose to click images from the app or use previously clicked ones. Tech Stack React-Native for building the app, python as the back-end, myScriptJS for client side OCR, tesseract-ocr and openCV for server-side OCR, and scikit for the ML stuff. Data Structures and Algorithms tesseract-ocr's and myScriptJS' engine for OCR and bag-of-words or KNN, whichever performs better, for classification. Shortcomings It can be implemented without much headache for high school level but more training data and a better researched machine learning model will be needed for supporting university level; which will happen in due time. However math extraction through OCR is still something that is not satisfactory, so it won't be supported. Future Scope Based on noteSort's core functionality, a full fledged platform can be built for sharing of class notes where students can upload whatever they think is important and it automatically categories itself. This can be also be incorporated in GeeksforGeeks. GFG, as of now, only caters to programming and computers in general and a typical student finds about it in college; usually when they are a sophomore/junior in college. But if noteSort is integrated with GFG, students will get a chance to know about GFG early. Not only will this give students access to GFG's quality content early, it will also make school students' transition to GFG easy. This has a great potential to solve the teaching problem that India has. Team - 1. Archit Mathur SKIT, Jaipur