EmotiPro

• 

  GUI made with tkinter to start and stop video stimulus

Inspiration: We have always been interested in facial recognition due to its diversity of applications in both security and for day-to-day uses in photography. One of our team members, Asher Lipman, attempted to design a methodology to detect changes in facial features in order to create a lie detector as part of his social science research project. When he told us about his research, we thought it was an amazing opportunity to combine both our interests and detect emotions using machine learning.

What it does: Our code detects the changes in emotions by landscaping coordinates as part of a pixelated image. By accessing the webcam on the user's computer, our software takes screenshots and analyzes the images to detect changes in emotions. We are also able to detect change's in the user's expressions when exposed to stimuli, such as a youtube video. The program will then be able to graph the changes in emotions over different portions of the video.

How we built it: We utilized the landscape identification and camera screenshotting properties code based off of the previous literature. Data for facial landscape locations were taken off of cohn-kanade facial expression database.

Challenges we ran into: It was difficult to install all the libraries needed to access the utilities allowing us to take screenshots on webcams and graphs.

Accomplishments that we're proud of: I learned how to draw a graph using matplot, and this was actually my first time coding with Python! It was also eye-opening to delve into the logistics behind the science of detecting facial expressions.

What we learned: We learned that python's syntax is simpler than java. I also learned the setup makes certain languages faster than others.

What's next for EmotiPro: We hope for this to be implemented either as a method to optimize marketing and analyzing consumer preferences or to assist in the

Built With

• matplotlib
• opencv
• python
• tkinter