traCEker | Devpost

traCEkr
Home Page
Shower tracker calendar where blue is days showered and red days missed.
Tracker where the user verifies their selfie to be inputted into the shower tracker.
Microsoft Azure trained model
MongoDB storing data

💭 Inspiration

Inspired by the constant complaints about the bad smell coming out of the ECE floor @ the E7 building and the DC building as a whole, we wanted to help change that. Not enough people in ECE and CS are spending time building positive hygiene habits, which is causing a ripple effect across the entirety of the Waterloo campus. To help combat this epidemic of bad body odor, we decided to create traCEker.

❓What it does

TraCEker is a platform where SE/ECE/CS students (but not exclusive to) can track how often they shower and how well they are doing relative to other users with a monthly leaderboard. By seeing other people's shower ranking, traCEker uses public humiliation to incentivize users to shower consistently. Users upload a selfie of them after they shower, which is detected by a classification vision model. Once the selfie is verified as a valid submission, the user is logged as showered. Users can use the calendar view to see how often they showered throughout the days, with showered days depicted in blue.

🛠 How we built it

We built traCEker using: -> React for the frontend website -> Python, Firebase, and Flask for the backend -> MongoDB to store all our data -> Trained Custom Vision using images to detect selfies with Microsoft Azure Custom Vision -> Figma to prototype the design

😕 Challenges we ran into

We ran into challenges in connecting the front end to the back end. All three of us were working individually on our tasks (front end, backend, and Microsoft Azure), and so we had to figure out at the end how to connect all three of these together.

😁 Accomplishments that we're proud of

We're proud that we were able to successfully identify selfies with high accuracy using what we trained using Microsoft Azure. Since we're all passionate about this subject matter of encouraging ECE/CS students to shower, we were proud that our method of verification worked and is relatively simple and easy to follow as a user.

🧠 What we learned

We learned how to consolidate all working parts of a project, from the front end, backend and the computer vision portion. By splitting up our tasks individually, we learned that we were a lot more efficient, which is especially important when time is limited.

💡What's next for traCEker

Next steps for traCEker would be to implement streaks to further incentivize consistency in showering. There will also be an option to add friends, as the ability to see your friends' progress or inability to shower will serve as social pressure.