DreamCatcher - YHack 2017

DreamCatcher is an image recognition program that uses machine learning to determine whether students are sleeping.

We scraped Google Images and manually took photos of people who are awake (engaged) and people who are asleep (sleeping) and fed our data into Clarifai's Image Recognition API to train our own model that detects if someone is sleeping or not.

We then created a website where you can put a link to an image or upload a local image, and the website will tell you the result of whether our model thinks the person is engaged or sleeping.

We also created a program that can take an image and select all the individual faces to crop them out.

Unfortunately, we did not have time to integrate both the face detecting algorithm and the model training. If we had more time, we could train our model using the cropped images so that it can better learn what an engaged and sleeping person looks like. Then, we can upload an image and use our face detecting algorithm to pick out the faces in the picture and be able to tell whether each person in the picture is engaged or sleeping, as well as an overall score of how many people are awake.

Share this project: