Date of Award

Spring 2025

Document Type

Thesis

Terms of Use

© 2025 Yihui Wu. This work is freely available courtesy of the author. It may be used under the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International (CC BY-NC-ND 4.0) license. For all other uses, please contact the copyright holder.

Degree Name

Bachelor of Arts

Department

Engineering Department

First Advisor

E. Carr Everbach

Abstract

SilentTalk is a noninvasive communication system that uses ultrasound technology to detect and interpret silent speech by analyzing lip and facial movements. This E90 project presents the design, construction, and experimental evaluation of a speaker–microphone device capable of emitting and capturing ultrasonic waves in the 20–55 kHz range. Data wereadS collected under five distinct conditions, including various static mouth gestures, and analyzed using Root Mean Square (RMS) smoothing and Fast Fourier Transform (FFT) techniques. The results show that the distinct features in the time and frequency domains correspond to different mouth shapes, supporting the feasibility of silent gesture classification. A method based on the standard deviation of the RMS voltage was proposed to detect the presence of a human face. The findings highlight the importance of consistent ultrasonic output and gesture variability in real-world applications. Future directions include improving hardware portability, improving signal stability, and integrating phonetic modeling and machine learning to enable silent real-time communication across languages.

Download

Included in

Engineering Commons

Share

COinS