Silent Speech Recognition Using AI and ML

Silent Speech Recognition Using AI and ML

K Swetha Sailaja*1, BANDARI ARUN*2, DANTURI ABHINAV*3, AKKENAPALLY CHARAN*4

*1 Assistant Professor of Department of CSE (AI & ML) Of ACE Engineering College, India.

*2,3,4 Students of Department of CSE (AI & ML) Of ACE Engineering College, India.

Abstract

Lip movement interpretation for speech recognition is a cutting-edge technology that enables communication without sound, benefiting individuals with speech and hearing impairments and enhancing human-computer interaction. This project focuses on Visual Speech Recognition (VSR) by analysing lip movements and converting them into accurate text using deep learning models. The system processes video frames, extracting key lip features using OpenCV, Dlib, and MediaPipe, and is trained on datasets such as the GRID Corpus and LRS2 to achieve robust recognition. The technology has real-world applications in communication aids, accessibility tools, and silent interfaces in sensitive environments.