-
-
Detects face other than primary user, captures zoomed in image of the face in the frame.
-
Displays a warning message
What it does
A face-detecting security measure that opens the webcam and takes a snapshot of a lurking individual's face during vulnerable moments such as logging into banking accounts, filling out tax forms, and registering for credit cards. Designed to protect sensitive information like SSN, bank account login information, etc. from unauthorized access. The primary function is to detect faces other than the user's face, if the person is looking directly at the screen from some distance. It activates whenever it detects the user is on a website/app where they may be sharing personal information. It will then display a warning message.
How we built it
Built using OpenCV webcam support and Python.
Challenges we ran into
We ran into issues surrounding face detection models that involved using training sets and creating predictive models, as the primary security function hinged upon the program being able to tell the difference between the primary user's face and someone else's face.
Accomplishments that we're proud of
We are happy about how much we learned and that we were able to achieve a webcam connection to the program.
What we learned
We learned about using virtual environments to test programs before actual deployment, a LOT of git and version control commands and troubleshooting, and about how facial detection actually works. We use it basically every day when we unlock our phones, and it was fascinating to learn about the different methods that are used to identify and classify facial features.
What's next for Face Off: Password Protection with Facial Detection
We want to add sophistication to the program so that it can recognize primary user facial features and store this information, so that it will not send warning messages due to two faces in the frame.
Presentation Link
Access our slides here
Log in or sign up for Devpost to join the conversation.