Affirmasaurus

Affirmasaurus Logo

Inspiration

We have been inspired by the spirit of the dino, aiming to boost moral and mental health of users, all whilst exploring new technology and ideas!

What it does

Solves the problem of mental health, tackles facial recognision of the user, creating a user-friendly gesture controlled quiz to supply dinosaur friends and motivational quotes ( quotes written by Caitlin Horton)

How we built it

To start, we used Open-CV python to perform facial recognition and detection of users from a camera. Upon pressing enter, the photo of the user (smiling) is stored in our directory, giving a touch of personalisation. A multiple choice quiz, implement using pygame with complex control and data flow, was shown to the user. This meant we had the opportunity to apply computational thinking methods like abstraction, decomposition, and use of object oriented paradigm. We then used conditional programming to select from a range of dinosaurs to greet the user. In addition, we randomly selected and read a line from a generated text file to act as a positive, uplifting and dinosaur themed motivational quote, displayed alongside the picture of the user. Rebecca Mewse created the majority of the handdrawn, artistic assets, including the overlay for the camera display and the dinosaur designs, which meant we have full, limitless control over the user interface, appearance and style. This meant that we ould create a calm, joyful space, reflecting our goal for this project.

Challenges we ran into

Causing great nerves, we managed to create a python file that was self-replicating beyond our intention and understanding (beyond our wildest nightmares). After careful inspection, Andrea Johnson realised that we had accidentally used a recursive subroutine (opening a new file), by misplacing a line of code. Here we truly realised and appreciated the power of programming as this one fatal line caused many trips to task manager. After removing this line of code, the program worked as intended. We encountered and overcome a bottlenecking issue, however not in the traditional way. This meant that the software was moving faster than the user could calmly comprehend. We had to meticulously revise our control structures to implement the appropriately placed and timed delays. This ensured that communication between various file was not distrupted. Because we were exploring new technology, we originally did not know how to link gesture controls on the camera display to the multiple choice answers. Through intensive research and team collaboration, we realised we could iteratively call a delay function, whilst importing functions between python files to dynamically send the direction of hand gestures to the quiz python script.

Accomplishments that we're proud of

Through encorporating a range of hand-drawn assets, we are proud to have implemented a consistent, out-of-the-box, dinosaur related theme. We truly feel this theme continues the uplifting ideas throughout this project and shows our creativity and consistency. In this project, we managed to create a fully functional computer vision model, including facial recognition and hand gesture recognition. This involved using new technologies and extensive research from Rebecca Mewse and Palak Shah, leading to our entire group feeling more confident after the great success with this model.

What we learned

This projected allowed us to fully explore different uses of github, especially within a team. This drastically improved our confidence and fluency with github, leading to us feeling like self-proclaimed, yet humble, "GitHub Gods". Following the theme of exploration, we explored a range of new technologies throughout this event. These included: facial recognition, hand gesture recognition, pygame, open_cv and threading. Some of our team members also used had to leanr python during this project.

What's next for Affirmasaurus

Make our source code into a local application Add a bank of questions that can be randomly selected to dynamically form the quiz Eye tracking software to control inputs such as "Enter" and multiple choice options. Faster execution achieved with GPUs