Deep Learning Strategy for Braille Character Recognition

Kausar, Tasleem; Sajjad, Muhammad; Kausar, Adeeba; Lu, Yun; Wasif, Muhammad; Ashraf, Muhammad Adnan

doi:10.1109/access.2021.3138240

Cited by 8 publications

(4 citation statements)

References 61 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The voice processing module converts extracted text to speech using a text-to-speech synthesizer (TTS) on a Raspberry Pi, amplified and played through a speaker. The system is built around a Raspberry Pi B+ processor board, which controls the entire setup [11].…”

Section: Related Workmentioning

confidence: 99%

Smart Book Reader for Visual Impairment Person Using IoT Device and Deep Learning

2024

IJSREM

View full text Add to dashboard Cite

People with normal vision can easily see the world around them and can read and write without difficulty. For those who are visually impaired, the Braille script enables them to read and write just like sighted individuals. According to WHO data from 2023, about 15 million people worldwide have significant vision loss. The Braille system uses cells with six raised dots, each dot numbered from one to six, arranged in two columns. This system is crucial for visually impaired individuals to keep up with the world around them. Providing Braille-assisted technology and incorporating it into daily life is essential to make life more comfortable and efficient for visually impaired people, enabling better communication with others. Key Words: Braille script, OBR.

show abstract

Section: Related Workmentioning

confidence: 99%

Smart Book Reader for Visual Impairment Person Using IoT Device and Deep Learning

2024

IJSREM

View full text Add to dashboard Cite

show abstract

“…Despite this challenge, D 2 NN, through continuous training and data updates, gradually enhances their recognition performance, bringing higher accuracy and the advantage of automated processing to the field of tactile perception recognition. 34 This progress opens up new possibilities for their application in the domain of tactile perception.…”

Section: ■ Introductionmentioning

confidence: 96%

“…This framework possesses advantages such as high speed, low power consumption, parallelism, and scalability, enabling it to efficiently accomplish large-scale computational tasks. , In recent years, the introduction of D 2 NN has provided a novel approach for image analysis, feature detection, object classification, and various domains requiring image and visual technology. − However, there is currently a relatively limited in-depth exploration regarding the direct application of D 2 NNs in the field of tactile perception. Despite this challenge, D 2 NN, through continuous training and data updates, gradually enhances their recognition performance, bringing higher accuracy and the advantage of automated processing to the field of tactile perception recognition . This progress opens up new possibilities for their application in the domain of tactile perception.…”

Section: Introductionmentioning

confidence: 99%

All-Optical Diffractive Deep Neural Networks Enabled Laser-Reduced Graphene Oxide Tactile Sensor for Braille Recognition

Liu,

Fang,

Zhang

et al. 2024

ACS Appl. Electron. Mater.

View full text Add to dashboard Cite

All-optical diffractive deep neural networks (D 2 NNs) show a wide range of applications in image recognition and artificial vision due to their advantages of high-speed parallel processing, low energy consumption, and excellent anti-interference ability. However, there is relatively limited research applying D 2 NNs for tactile perception. In this study, we propose an automatic Braille recognition method based on D 2 NNs and tactile sensors. A flexible molybdenum disulfide-doped laser-reduced graphene oxide (LRGO/MoS 2 ) tactile sensor was fabricated with the laser direct writing method. The LRGO/MoS 2 tactile sensor shows a sensitivity of 9.8 kPa −1 , with a response/recovery time of 0.14/0.10 s and excellent cyclic stability. The tactile sensor can be employed to capture Braille character information in real time and convert it into digital signals as inputs for all-optical D 2 NNs. The automatic recognition of Braille characters is achieved in the all-optical D 2 NNs with five diffraction layers, and the system finally can realize a recognition accuracy of 100% for Braille recognition. The strategy of integrating flexible tactile sensors with all-optical deep learning paves a path for realizing a low-cost, fast, accurate, and efficient tactile recognition system.

show abstract

“…It is necessary to train the hands sensitivity to the raised dots and understand and memorize their combination in forming a letter. Braille classification is essential to comprehend braille letters because braille letters are different from the regular alphabet (Herlambang et al, 2021;Hossain et al, 2018;Kausar et al, 2021).…”

Section: Introductionmentioning

confidence: 99%

Utilizing Convolutional Neural Network for Learning Web-Based Braille Letter Classification System

Ridwan,

Purbolingga,

Hanisah

2024

CNAHPC

View full text Add to dashboard Cite

This paper aims to facilitate prospective teachers and people who want to learn braille letters. The system designed is a website that will classify braille letters using the convolutional neural network (CNN) method with the activation functions used, namely ReLU and Softmax. In this research, the input is an image of braille letters with grayscale elements. The output of the data is a regular alphabet letter. Most of this research data consists of training and testing data, which is 2,722 pieces. The accuracy results obtained in the data training process using Max Pooling and epoch 30 for data is 92.15%, epoch 50 is 94.58%, and for training data with epoch 100 is 96.64%. The test results using the system produce an accuracy value of all braille letter image data of 92.30%. Furthermore, for better system development, it is recommended to use hyperparameter tuning to minimize classification uncertainty in braille letter images.

show abstract

Deep Learning Strategy for Braille Character Recognition

Abstract: This research is supported in part by the joint fund of basic and applied basic research fund of Guangdong province under grant number 2020A1515110498.

Cited by 8 publications

References 61 publications

Smart Book Reader for Visual Impairment Person Using IoT Device and Deep Learning

Smart Book Reader for Visual Impairment Person Using IoT Device and Deep Learning

All-Optical Diffractive Deep Neural Networks Enabled Laser-Reduced Graphene Oxide Tactile Sensor for Braille Recognition

Utilizing Convolutional Neural Network for Learning Web-Based Braille Letter Classification System

Contact Info

Product

Resources

About