Built a dataset of Gujarati Isolated Handwritten Characters and Recognition through deep learning
Main Article Content
Abstract
In the current era with the rise of new machine learning algorithms, particularly deep learning, the demand for large, high-quality datasets has grown significantly, especially in handwritten character recognition (HCR). While several Indian languages have publicly available benchmark datasets, a few, including Gujarati, still lack such resources. This paper addresses an attempt to build a dataset for Gujarati isolated handwritten characters and to recognize the isolated Gujarati handwritten vowels and consonants. The dataset is collected from 692 writers of varying ages, genders, qualifications, and professions. The dataset consists of 63,664 samples for 46 classes including 34 consonants and 12 vowels where 1384 images of each character. The proposed model was run with an 80:20 training and testing ratio, using 7, 10, 20, 30, & 40 epochs. The model showed promising results and achieved the highest training accuracy 90.92%, and the highest testing accuracy 89.51%.
Downloads
Article Details
COPYRIGHT
Submission of a manuscript implies: that the work described has not been published before, that it is not under consideration for publication elsewhere; that if and when the manuscript is accepted for publication, the authors agree to automatic transfer of the copyright to the publisher.
Authors who publish with this journal agree to the following terms:
- Authors retain copyright and grant the journal right of first publication with the work simultaneously licensed under a Creative Commons Attribution License that allows others to share the work with an acknowledgment of the work's authorship and initial publication in this journal.
- Authors are able to enter into separate, additional contractual arrangements for the non-exclusive distribution of the journal's published version of the work (e.g., post it to an institutional repository or publish it in a book), with an acknowledgment of its initial publication in this journal.
- Authors are permitted and encouraged to post their work online (e.g., in institutional repositories or on their website) prior to and during the submission process, as it can lead to productive exchanges, as well as earlier and greater citation of published work
- The journal allows the author(s) to retain publishing rights without restrictions.
- The journal allows the author(s) to hold the copyright without restrictions.