Inspiration
The app was inspired by our own studying and stress release habits. We noted the ways we relieved short term stress to focus on our work. We know that a stress free mind leads to better knowledge retention..
What it does
The program analyzes the facial patterns of the user as they work/study. Using machine learning, it recognizes stress then determines if there is prolonged stress. If there is prolonged stress the user receives a message prompting various activities to relive stress, then get back at it.
How we built it
We took pictures with different facial expressions. Then used the machine learning protocol to recognize a stressed face and a relaxed face. We implemented an algorithm to run the detection in the background and prompt the user when the a stressed face is detected for a prolonged threshold.
Challenges we ran into
We set out to find a repository of many facial expressions to use to analyze any user. We found that the best available repository, does not contain black people, and would not work on our face. Instead of taking thousands of samples we used just our face for the application.
We had technical issues with one of our team members computers battery draining, and charger shorting.
Accomplishments that we're proud of
We are proud to have completed our planned project within the allotted time. We are glad to have successfully used machine learning to learn our facial expressions.
What we learned
We learned how to make a machine learning protocol more accessible to a variety of people.
What's next for sfs
We will expand the repository to work for everyone by implementing the available repository of faces with more black faces. This will make it accessible to more people. We may interface our program with personal health product to detect stress in other ways. This may include smart watches that can detect the heart rate.
Log in or sign up for Devpost to join the conversation.