Handwritten Gujarati Character Recognition Using Machine Learning and Deep Learning
Yogiraj Zala, Krishn Limbachiya, Ankit K. Sharma, Pooja Shah · Advances in computer science research · 2023
Gujarati character recognition has always been a field of study in need of improved techniques to reach a satisfactory degree of accuracy.This study developed a technique for predicting handwritten Gujarati numerals using various feature extraction and classification techniques.Deep learning models, including MobileNet, DenseNet, and NasNetMobile, are components of the feature extraction technique.XGBoost, Random Forest, Logistic Regression, Naive Bayes, Stochastic Gradient Descent, Decision Tree, and K-Nearest Neighbor are some of the techniques used to classify Gujarati numerals.This study used over 14,000 images of Gujarati numerals for an experiment and found significant outcomes.