Gesture Guard

Inspiration

One of our members had been playing around with image recognition libraries in recently and we were also interested in exploring security-based projects to learn more about cybersecurity concepts which is how we ended up with this project idea

What it does

In a similar vein to FaceID, Gesture Guard allows users to create a passcode and login with their physical features, specifically, using different hand symbols! Users can set up a sequence of hand symbols, and then display those symbols to a camera to login without authenticating through the classic method (typing)

How we built it

Our project was a full-stack Python project. It was built utilizing several libraries, mainly MediaPipe for the hand symbol recognition, Flask for the backend, and PyQT5 for the frontend. MongoDB was used for our database.

Challenges we ran into

Learning new technologies :)
Integrating each component of the app
Learning how to use MongoDB and Flask
Testing the user flow + HTTP responses

Accomplishments that we're proud of

Finishing the project :)

What's next for Gesture Guard

Currently, we are limited to a fixed number of pre-defined hand symbols. In the future, we think it would be ideal to empower users to create their own hand symbols to use, and also increase the overall security of the program. If given more time, we would look into applying machine learning technology to support this. As this is made for security purposes, we would also look into bolstering our encryption process. For example, we intended to encrypt the registration data with public/private key pairs, but unfortunately unable to due to time constraints. Additionally, with more time we would take more steps to ensure the robustness of our application through improvements in error handling and with a regression test suite.