Neuro-Fuzzy Based Voice Enabled Smart e-Learning System for Disabled and Visually Impaired Students

Abstract

E-learning has now grown as an essential means of information and skills acquisition for learners. Visually disabled individuals, on the other hand, have little or limited or no access towards this tool due to the lack of a user interface that is accessible to them. The Voice-enabled e-Learning approach might be able to help with this problem. The aim of evolving such an approach or system is to help visually impaired and permanently disabled students learn a desired topic through the system in a comfortable manner on utilizing the voice instructions. We suggest a Neuro-Fuzzy Based Voice-Enabled Smart e-Learning (NFVSe-L) Model after undertaking a historical study and determining the use of the latest and best e-Learning technologies especially when considering the characteristics of disabled and visually impaired students. This system involves four significant subdivisions: Student Verification, Speech Recognition, Instruction Recognition, and Information Retrieval subdivisions. Preprocessing of the retrieved speech signal, Linear Prediction Cepstral Coefficients (LPCC) for feature extraction and neuro-fuzzy methodology for synthesizing and understanding the voice signal are all utilized in the Speech Recognition Subdivision. Pattern matching quantization is used in the Instruction identification subdivision. The experimental findings indicate 5.73% of minimal mean word error rate in student verification subdivisions and 97.7% of mean accuracy in speech recognition subdivisions.