Next Article in Journal
Analyzing User Digital Emotions from a Holy versus Non-Pilgrimage City in Saudi Arabia on Twitter Platform
Next Article in Special Issue
Multilateration Approach for Wide Range Visible Light Indoor Positioning System Using Mobile CMOS Image Sensor
Previous Article in Journal
Optimization of Computed Tomography Angiography Protocols for Follow-Up Type B Aortic Dissection Patients by Using 3D Printed Model
 
 
Article

BengaliNet: A Low-Cost Novel Convolutional Neural Network for Bengali Handwritten Characters Recognition

1
Department of Computer Science & Engineering, Rajshahi University of Engineering & Technology, Rajshahi 6204, Bangladesh
2
School of Computer Science and Engineering, The University of Aizu, Aizuwakamatsu, Fukushima 965-8580, Japan
*
Author to whom correspondence should be addressed.
Academic Editor: Seokwon Yeom
Appl. Sci. 2021, 11(15), 6845; https://doi.org/10.3390/app11156845
Received: 14 June 2021 / Revised: 14 July 2021 / Accepted: 23 July 2021 / Published: 25 July 2021
(This article belongs to the Special Issue Intelligent Processing on Image and Optical Information, Volume III)
As it is the seventh most-spoken language and fifth most-spoken native language in the world, the domain of Bengali handwritten character recognition has fascinated researchers for decades. Although other popular languages i.e., English, Chinese, Hindi, Spanish, etc. have received many contributions in the area of handwritten character recognition, Bengali has not received many noteworthy contributions in this domain because of the complex curvatures and similar writing fashions of Bengali characters. Previously, studies were conducted by using different approaches based on traditional learning, and deep learning. In this research, we proposed a low-cost novel convolutional neural network architecture for the recognition of Bengali characters with only 2.24 to 2.43 million parameters based on the number of output classes. We considered 8 different formations of CMATERdb datasets based on previous studies for the training phase. With experimental analysis, we showed that our proposed system outperformed previous works by a noteworthy margin for all 8 datasets. Moreover, we tested our trained models on other available Bengali characters datasets such as Ekush, BanglaLekha, and NumtaDB datasets. Our proposed architecture achieved 96–99% overall accuracies for these datasets as well. We believe our contributions will be beneficial for developing an automated high-performance recognition tool for Bengali handwritten characters. View Full-Text
Keywords: banglalekha dataset; bengali handwritten character recognition; CMATERdb dataset; computer vision; dropout; ekush dataset; low-cost convolutional neural network; NumtaDB dataset; pattern recognition; supervised learning banglalekha dataset; bengali handwritten character recognition; CMATERdb dataset; computer vision; dropout; ekush dataset; low-cost convolutional neural network; NumtaDB dataset; pattern recognition; supervised learning
Show Figures

Figure 1

MDPI and ACS Style

Sayeed, A.; Shin, J.; Hasan, M.A.M.; Srizon, A.Y.; Hasan, M.M. BengaliNet: A Low-Cost Novel Convolutional Neural Network for Bengali Handwritten Characters Recognition. Appl. Sci. 2021, 11, 6845. https://doi.org/10.3390/app11156845

AMA Style

Sayeed A, Shin J, Hasan MAM, Srizon AY, Hasan MM. BengaliNet: A Low-Cost Novel Convolutional Neural Network for Bengali Handwritten Characters Recognition. Applied Sciences. 2021; 11(15):6845. https://doi.org/10.3390/app11156845

Chicago/Turabian Style

Sayeed, Abu, Jungpil Shin, Md. Al Mehedi Hasan, Azmain Yakin Srizon, and Md. Mehedi Hasan. 2021. "BengaliNet: A Low-Cost Novel Convolutional Neural Network for Bengali Handwritten Characters Recognition" Applied Sciences 11, no. 15: 6845. https://doi.org/10.3390/app11156845

Find Other Styles
Note that from the first issue of 2016, MDPI journals use article numbers instead of page numbers. See further details here.

Article Access Map by Country/Region

1
Back to TopTop