Enhancing STEM Education Accessibility for Blind and Low Vision Students

This study focuses on enhancing STEM education accessibility for blind and low vision students through the development of innovative assistive technologies. The primary challenge addressed is the limited engagement and effectiveness of traditional teaching methods for visually impaired students, particularly in science and mathematics. Leveraging advancements in computer vision, machine learning, and internet of things technologies, the proposed solution includes a mobile application integrated with YOLObased object detection and real-time audio feedback. Methods employed involve capturing and preprocessing images of geometric shapes, educational materials, and currency, followed by rigorous model training using Google Colab. The system’s architecture incorporates YOLOv8 for real-time object detection, combined with text-to-speech and speech-to-stext technologies to provide comprehensive audio descriptions and interactive feedback. Results demonstrate significant improvements in detection accuracy and user engagement, with YOLOv8 achieving high precision in object recognition tasks. Performance standards were evaluated through confusion matrices, highlighting the model’s robustness across various educational scenarios. The developed system effectively enhances independent learning and engagement among visually impaired students, promoting inclusivity and equity in STEM education.