You need to type in your password to access almost any sensitive aspect of your life. Whether you want to see your emails, social media accounts, or academic portals, the login page is the gate keeper for all of them.

This makes our passwords all the more important to safeguard because with it anyone can gain access to any of our accounts. Thus, what we propose is to use our webcam, which most devices (laptops and smartphones) have, to track the movement of our eyes. This allows us to trace out patterns with our eyes that we can map to the various passwords we have for our different accounts for different services. By combining the webcam's images with openCV, we can decode the pattern and check if it maps to a password which we can then insert into the password input box of the form for the user. In the case that the pattern is incorrect we can insert an arbitrary password into the password field.

We have developed this hack for web browsers first, but strongly believe that the same technologies could be extended to be used on mobile devices as well. This way the user will have the same exact user experience when logging into a service on the phones and on the laptops. We strongly believe that this product will help bridge the ease-of-use disparity between laptops (where it's easy to type) and mobile devices (where it is a little bit harder).

As for the group of users we target, it is everyone who could potentially read this (meaning anyone with an electronic device with internet access). Our product, Traze, can make it easier for anyone and everyone to login in to anything!

The coolest part of the project in our opinion is how we use openCV to decode the pattern created by our eyes. The eyes cover so little real estate on the screen that is makes it not only a hard thing to track, but also a more secure method for protecting our passwords.

But is Trace really that cool? You decide :)

Share this project: