Browsing by Author "Neranga, K.T."

Filter results by typing the first few letters
Now showing 1 - 1 of 1
  • Thumbnail Image
    PublicationEmbargo
    Mobile and Simulation-based Approach to reduce the Dyslexia with children Learning Disabilities
    (Institute of Electrical and Electronics Engineers, 2022-09-16) Muthumal, S.A.D.M; Neranga, K.T.; Harshanath, S.M.B; Sandeepa, V.D.R.P; Lihinikaduwa, D.N.R; Rajapaksha, U.U.S.K
    Learning disabilties are frequently overlooked while treating various limitations. The primary causes for such a dropout might be a lack of awareness of the risks and access to adequate medical care. Because Learning Disabilities are neurological illnesses, their etiology is unknown. Learning Disabilities do not have a therapy or a cure. Dyslexia, Dysgraphia and Dyscalculia are the most common types of Learning Dis-abilities among school children. There are numerous approaches for diagnosing and treating this ailment available today. The optimal method is to make use of a mobile application. Existing applications either do not adequately handle the problem or have minor limitations in terms of meeting genuine expectations. We can mitigate this dysfunction by utilizing mobile and simulation-based technologies. Furthermore, earlier research has not given contact and curiosity among children a significant emphasis and children have not interacted with robots. Therefore, this paper, introduces interactive and collaborative mobile appli-cation called 'Helply' with a robotic based simulation that may foster learning and help children improve and encour-age Color identification skills, Reading skills and Short-term memory skills the learning process while reducing their other dyslexic disorders. Also, NAO Robot is used to taking inputs as voice clip using voice recognition technology and images using image processing technology which embedded in NAO robot. We constructed three models for three distinct disorders and obtained accurate findings. CNN model for colour disorder had a training accuracy of 93%, FFNN model for short-term memory disorder had a training accuracy of 98%, and CNN model for reading disorder had a training accuracy of 94%.