Developing a Speech Recognition System for Educational Settings

The field of education is not an exception to the revolutionary developments brought about by the swift integration of Artificial Intelligence and Machine Learning. In this work, we propose a Digital Twin to Assist Teaching Learning (DTATL) model, a ground-breaking method intended to automate and improve the teaching-learning process using intelligent speaker recognition, in response to the rapidly growing trend of digital learning. The main goal of our research project is to use the Support Vector Classification (SVC) model to construct a speaker recognition system. This concept is incorporated into an approachable Flask-based user interface, making it available to both instructors and students. We used the Mel-frequency Cepstral Coefficients (MFCC) for the feature extraction technique and achieved remarkable results by utilizing the enormous potential of machine learning, with speaker identification training, and testing accuracies of 95%, and 84% respectively. This study also offers deep insights into the application and impact of the DTATL in the educational sector by offering a thorough overview of its creation, results, and prospective advantages.