Open AccessYORÙBÁNET: A DEEP CONVOLUTIONAL NEURAL NETWORK DESIGN FOR YORÙBÁ ALPHABETS RECOGNITION
Author(s) -
Oyeniran Oluwashina Akinloye,
Oyebode Ebenezer Olukunle
Publication year - 2021
Publication title -
international journal of engineering applied science and technology
Language(s) - English
Resource type - Journals
ISSN - 2455-2143
DOI - 10.33564/ijeast.2021.v05i11.008
Subject(s) - epoch (astronomy) , convolutional neural network , computer science , test set , character (mathematics) , artificial neural network , set (abstract data type) , matlab , artificial intelligence , deep learning , value (mathematics) , pattern recognition (psychology) , speech recognition , machine learning , mathematics , computer vision , stars , geometry , programming language , operating system
Numerous works have been proposedand implemented in computerization of varioushuman languages, nevertheless, miniscule efforthave also been made so as to put YorùbáHandwritten Character on the map of OpticalCharacter Recognition. This study presents a noveltechnique in the development of Yorùbá alphabetsrecognition system through the use of deeplearning. The developed model was implementedon Matlab R2018a environment using thedeveloped framework where 10,500 samples ofdataset were for training and 2100 samples wereused for testing. The training of the developedmodel was conducted using 30 Epoch, at 164iteration per epoch while the total iteration is 4920iterations. Also, the training period was estimatedto 11296 minutes 41 seconds. The model yielded thenetwork accuracy of 100% while the accuracy ofthe test set is 97.97%, with F1 score of 0.9800,Precision of 0.9803 and Recall value of 0.9797.