MUMBAI, India, June 30 -- Intellectual Property India has published a patent application (202641056291 A) filed by Rajalakshmi Engineering College on May 04, 2026, for Real-Time Captioning And Lip Reading Assistant For Hearing Impaired Learners.
Inventors include Dr. M. Thangamani; and Mrs. Kavitha V Kakade.
The application for the patent was published on June 26, 2026, under issue no. 26/2026.
Abstract: This invention introduces an assistive system aimed at supporting hearing-impaired learners in understanding spoken communication more effectively. It addresses the shortcomings of conventional captioning tools, particularly in noisy conditions or when speech is unclear. The system simultaneously captures audio and video of the speaker, with emphasis on lip and mouth movements. It produces two text outputs: one from Automatic Speech Recognition (ASR) based on audio,, and 370 another from a deep learning-driven Visual Speech Recognition (VSR) module that interprets lip movements. These outputs are merged using a confidence-based fusion method to generate a more accurate transcript. Additionally, a Convolutional Neural Network (CNN) is used to detect facial expressions, providing emotional context. The resulting captions are displayed through a customizable interface compatible with various devices, ensuring improved accessibility and learning effectiveness.
Disclaimer: Curated by HT Syndication.
