Design and Implementation of a Voice-Based Electronic Results System for Physically Challenged Students

Abstract

The growing population of physically challenged students in academic institutions necessitates the development of inclusive digital solutions. This research focuses on designing and implementing a voice-based electronic results system for Adekunle Ajasin University’s existing result processing system, AVERS (Ajasin Varsity Examination Result System). The proposed system leverages Artificial Neural Networks (ANN) for speech recognition and automation and integrates Hidden Markov Models (HMMs) for sequential speech processing to provide hands-free operations for students, particularly those with dyslexia and other physical impairments. Unlike conventional result processing systems, this solution uniquely combines ANN and HMM in a hybrid architecture tailored specifically for academic environments, ensuring accurate speech recognition even in noisy university settings and accommodating diverse speech patterns among users. This adaptive approach enhances system robustness and user-friendliness, making it distinct from general-purpose voice systems such as Google Voice. This paper presents the problem, system design, and methodology, culminating in a comprehensive evaluation of the voice-enabled solution. The findings, which show a commendable accuracy rate of 92% in recognizing voice commands and an average response time of 2 seconds, demonstrate that incorporating voice automation significantly enhances the inclusiveness and usability of the institution’s result-processing system, fostering independence, privacy, and improved accessibility for physically challenged students.