Highlighting the power of machine learning in creating inclusive solutions that cater to diverse user needs.

 

Conference paper

The system presented in this project aims to Break down communication barriers for visually impaired individuals by translating written Braille into both text and audio formats. It leverages advanced machine learning techniques to accurately interpret Braille symbols and convert them into readable text. The system employs convolutional neural networks (CNNs) to improve the precision of Braille recognition. In addition, text-to-speech functionality is integrated to provide audio output, making the system more accessible. This project highlights the power of machine learning in creating inclusive solutions that cater to diverse user needs.

Braille Recognition Using Convolutional Neural Network.



Comments

Post a Comment

Popular posts from this blog

How to type a text using braille directly on your iPhone with Braille Screen Input?

Image
Discover how to type using braille directly on your iPhone with Braille Screen Input , an innovative accessibility feature designed for users who are blind or visually impaired. This workshop will guide you through how to set it up, Understand basic finger gestures , and start typing efficiently using braille on a touchscreen . The session will include a presentation with demonstrations , followed by a Q&A for personalized support and discussion. Workshop entitled '' Understanding Braille Screen Input on iPhone ''. Wednesday, January 28, 2026 11:00 AM to 12:00 PM PT. Online / San Diego.
Read more

The development of braille display utilizing rotating braille-embossing disks with AI-based character recognition.

Image
In this study , we have developed a hardware-optimized rotating disk-based Braille display . It requires a fewer actuators compared to the commercially available Braille notetakers that rely on piezoelectric actuators or electromagnets. An evaluation system was also constructed to quantify the accuracy and response time of the Braille output process using this newly developed display . The Braille display utilizes a Braille-based disk structure that indirectly represents braille patterns. Experimental results demonstrated that the rotational accuracy of the disk was within 2.7°, and a bidirectional rotation reduced the rotation pulse up to 200 compared to unidirectional rotation. Furthermore, the mean output time for a set of ten frequently used English words was measured to be 1600 ms. When YOLO V8-n AI (artificial intelligence) library was used to detect the characters and the position of a user’s finger model, the system consistently exhibited an average confidence level of ∼86 %...
Read more

From Braille to Artificial Intelligence: Shaping a more accessible future.

Image
UNRIC spoke with Lars Bosselmann, the Executive Director of the European Blind Union , about the role of technology in improving accessibility and efforts to secure the rights of visually impaired people . What is the estimated number of people who are blind or partially sighted in Europe? Could you explain the difference? It is estimated that over 30 million people are either blind or partially sighted in Europe, including countries outside the European Union. In general, blind people have no vision, while partially sighted people retain some vision, from light perception to the ability to see shapes, colours, or contrasts. The definitions of blindness and low vision vary across countries, which can impact these estimates. In your view, what are the biggest obstacles for visually impaired people in our society? Apart from the physical barriers, one of the most significant challenges is the accessibility of information and communication technologies. When technology is not inclusive, i...
Read more