Virtual Assistant using Silent Speech Recognition for Speech Impairments Assistance
Summary
No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.
Summarise with VIRAT Free to try with your Science Catalogs account.Abstract
Silent Speech Recognition (SSR) provides a promising approach for developing virtual assistants that support individuals with speech impairments. SSR systems interpret speech from non-acoustic cues such as lip movements, facial electromyograms (fEMG), or tongue motion, helping to overcome limitations of conventional voicebased interfaces that perform poorly in noisy environments or for users unable to vocalize clearly. This work presents a vision-based SSR system that enables silent interaction with a virtual assistant using only webcam-captured lip movements. The prototype eliminates the need for audio input or external sensors and implements an end-to-end pipeline for lipregion detection, CNN-LSTM-based classification, and real-time command execution. Designed specifically for accessibility, the system provides a silent, contactless, and user-friendly communication method for individuals with speech impairments.
Fields of Study & Subjects
Speech and Audio Processing
Hand Gesture Recognition Systems
Gaze Tracking and Assistive Technology
Publication details
- Journal / source
- International Conference on Intelligent Systems for Communication, IoT and Security (ICISCoIS)
- Pages
- 1–7
- Publication date
- 8 January 2026
- Language
- EN