BRAILLAI: A Two Way Web-Based Braille Translation Using Convolutional Neural Network with Gamified E-Learning System Applied for Moderately Visually Impaired and Sighted Individuals

This research introduces a web-based platform designed that utilizes convolutional neural networks along with a gamified e-learning approach. Aimed at both moderately visually impaired and sighted users, The platform leverages sophisticated Convolutional Neural Network (CNN) models to effectively translate text into Braille and vice versa, facilitating smooth communication among users with varying visual abilities. This research addresses BRAILLAI's effectiveness in enhancing Braille understanding and its capacity to connect visually impaired and sighted communities, thereby contributing to a more inclusive society. The project underwent testing in accordance with ISO 25010. The results by phase of the study test results demonstrated high levels of functional suitability, performance efficiency, reliability, and usability, as outlined by the ISO model. Moderately Visually Impaired and Sighted Individuals rated the functional suitability, performance efficiency, reliability and usability with means ranging from 4.9 to 4.49, indicating a “Very Good” level of quality, affirming the application's adherence to ISO standards and requirements. These findings underscore the positive performance of the website across all evaluated criteria, showcasing its effectiveness in meeting the needs of both Moderately Visually Impaired and the Sighted individuals. Furthermore, the findings indicate that the integration of computer vision technology with educational content delivery can provide effective learning experiences for diverse learners in language learning.