For the Machine Learning course, students were asked to explore and prototype new interactions that could help users change their daily activities with a bit of ‘artificial’ help. After multiple ideation sessions, the team got interested in the question: how might we create awareness of tics and habits to motivate users to change behaviors during working from home scenarios?

TicTrak is a customizable real-time tracking system for tics and habits in front of their laptops, based on the team’s approach that tracking behaviors and generating reports will encourage users to change these habits. The user interface includes two main interactions: logging and training behaviors that the user is willing to track and a display that counts behavior repetitions in real-time.

The ML process behind this software includes three steps: feeding the system with pictures that are labelled with tags, training to start guessing and constantly looking for and counting similarities in live video.