The Handwritten Devanagari Character Recognition Using Deep Learning

Abstract

English has 372.9 million first-language speakers. Handwritten English character recognition using Latin script has been extensively researched. Devanagari script is used to write multiple languages of South Asia with over 532.1 million speakers. Research on Devanagari handwritten character recognition has been limited. This paper explores Devanagari handwritten character recognition using deep convolutional neural networks based on LeNet-5.