Get rid of your Tcard, pay with your face


Forgetting to carry your Tcard and having to go all the way back to your residence to get it back is the most tedious part of being at UofT(except of course midterms). Eliminating the need for this has been our source of inspiration and motivation. Also, some TCards have not been working at some places for some students around campus.

What it does🔎

Our project enables students to securely authenticate payments on campus using their face. The registration takes less than a minute, our database securely stores this data and links it to the student number. The ability to add multi factor authentication in the future will further improve our security.

How we built it🔨

Our facial recognition model is created by transfer learning using MobileNet with over 3000 images. We have used TensorFlow and Keras to enable us to run our model. This model was then integrated using Django into a fully functional website, a secure database SQL was built to store and use the data to authenticate payments. Even though our project currently runs locally, with slight tweaks it has the potential to run at a much larger level too. The aesthetic design and easy-to-use property of our final product will enable a smooth transition into tPay.

Challenges we ran into⚠️

Training and getting our face recognition model to work was one of the trickier tasks involved. We had to use multiple databases to train our model to recognise random faces and backgrounds as failed cases.

Accomplishments that we’re proud of🥇

After multiple attempts and thousands of training images, we were finally able to produce a full function face detection bases payment authenticator. Our model not only identifies the person in front of the camera, it is also able to decline payments if a non-student shows up to authenticate. Our website is pleasing to see and effortless to operate.

What we learned🧠

None of our project members was familiar with all the frameworks used, everyone learned at least one new model to use in the fields of computer vision and web development. All our project members are first year students at UofT, but our combined experience helped us complete this project in the allotted time(thanks to daylight savings too!). Our belief in teamwork also increased during the course of this hackathon. Most of all, it solidified our view that copious amounts of clean data is what drives the best artificial intelligence systems.

What’s next for Weed Detech💭

With a bit more of computing power, slight optimisation and automation tPay is completely ready to go into testing phase and eventually be applied so that students have one less thing to worry about at university.


  • Shivesh Prakash
  • Rishit Dagli
  • Sai Manish
  • Alex Rosen

Check it out here: GitHub Repo