Abstract

Handwriting recognition is the ability of a computer to receive and interpret intelligible handwritten input from sources such as paper documents, photographs, touch-screens and other devices. The purpose of the research is to design develops, construct, deploy, and test a convolutional neural network (CNN) for handwriting recognition. The CNN for handwriting recognition was developed with Python and MNIST dataset was used. CNN was evaluated together with Simple Neural Network (SNN) and it was found that a CNN operating on well-tuned hardware with GPU and adequate training data can recognize numbers with an accuracy of up to 98.7 percent. The accuracy and speed of the model can be improved by expanding the dataset, increasing the number of epoch runs, and executing it on parallel hardware. Using these strategies, an accuracy of up to 99.89 percent, can be achieved. Keywords: CNN, SNN, Tensor Flow, Neural Network Architecture, Confusion matrix DOI: 10.7176/CEIS/14-1-05 Publication date: February 28 th 2023

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call