Presenter(s)
Ishwar Sandip Jadhav
Files
Download Project (885 KB)
Description
Do you know the feeling when you go out and realize you left your phone at home? Yes, this feeling is uneasy. Our mobile device these days are more than just a tool for communication, it has a lot of serious data and personal information, such as contacts, emails, photos, and passwords that can put our privacy at risk. A modern security feature like FaceID though prominent is not sufficient to safeguard our data. “For instance, Security researchers attending the annual Black Hat hacker convention in Las Vegas have managed to bypass the iPhone FaceID user authentication in just 120 seconds. These researchers were able to demonstrate that they could bypass the FaceID user authentication and access the iPhone of the victim in less than 120 seconds. To do so, they needed three things: a pair of spectacles, some tape, and a sleeping or unconscious iPhone user”. Therefore, in this project we would like to develop a secure user authentication via facial expression analysis.II. Methods Using facial landmark point detection, facial gesture recognition involves collecting a dataset of facial gesture images or videos, pre-processing the data by identifying and extracting facial landmarks, and extracting features for recognition. A machine learning model then is trained, its performance is evaluated, it is integrated into an authentication system, potential security risks have been considered, and user testing is done. The facial landmark detection and feature extraction method such as MediaPipe is effective and reliable. Key facial features are identified and extracted as part of the approach to produce relevant features for recognition. These characteristics are used to train a machine-learning model to differentiate between genuine and fake facial motions. The user is prompted to make a particular gesture, which the system compares to the trained model to confirm their identity, spoof-blocking methods and potential security theft.III. Significance When paired with facial recognition technology, facial gesture recognition is a promising technique that can offer additional benefits for unlocking mobile devices. Asking the user to make a specific facial gesture in addition to face recognition, adds an extra layer of security and can assist prevent unauthorized access to the device. Also, instead of inputting passcodes or utilizing fingerprint scanners, face movements can unlock a smartphone more quickly and conveniently. Facial gesture recognition can also increase accessibility for people who find entering a passcode or using a fingerprint scanner challenging. Last but not least, it can provide a pleasant and unique user experience by letting users select memorable or meaningful actions that increase a sense of ownership.
Publication Date
4-19-2023
Project Designation
Course Project 202310 CPS 595 P1
Primary Advisor
Van Nguyen
Primary Advisor's Department
Computer Science
Keywords
Stander Symposium, College of Arts and Sciences
Institutional Learning Goals
Scholarship; Faith; Traditions
Recommended Citation
"Facial Gesture Recognition for User Authentication" (2023). Stander Symposium Projects. 3210.
https://ecommons.udayton.edu/stander_posters/3210
![Facial Gesture Recognition for User Authentication](https://ecommons.udayton.edu/stander_posters/4211/thumbnail.jpg)
Comments
Presentation: 10:45 a.m.-12:00 p.m., Kennedy Union Ballroom