We live in a time were everything lures for our attention, depression and loneliness has been an increasing problem. What if we use deep learning to help reduce te problem?
What it does
It shows a random meme to the user and captures the facial expresion to check the current emotion a classifier built with TF 2.0. It records the emotion at that given time, and writes a log file. This would help in a future to create a recurrent neural network model to show the images to the user that will increase their mood.
How I built it
I used the kaggle dataset Emotion and identity detection from face images to create an emotion classifier
I used my previous project Meme_Classification_Tensorflow to have a bunch of images
I capture the image from the webcam with opencv
I created the app with kivy (I would like it to be multiplatform later)
Challenges I ran into
The biggest problem was the idea! I regret that It came one week before the hackathon ends, after that I had to learn kivy I had not used it before.
Accomplishments that I'm proud of
Finishing on time and learning new python tricks!
What I learned
What's next for cheer(me)
improve the app! make it multiplatform, create the rnn based on the log files