Machine Learning Symbol Recognition
No Thumbnail Available
Authors
Stefánsson, Birgir
Date of Issue
2023-04-28
Type
Presentation
Language
Subject Keywords
Other Titles
Abstract
This Machine Learning Symbol Recognition project uses machine learning to train models to recognize what symbol the user draws on their screen. The process of being able to identify these symbols consists of generating thousands of symbols using Python, training convolutional neural networks, and accepting user input through a website. The symbols I chose to use were both self-determined and known symbols. Once the user draws a symbol on the website and hits submit, the symbol will be resized and fed to that model. The model or process of the neural network will analyze the image grid by grid trying to determine what symbol was drawn.