The biggest problem with online learning is a lack of face to face human contact, and difficulty reading emotional state of the students. We bring this human contact back by including an interactive emotional component and a collaborative environment. We can detect 7 emotions of students: happiness, sadness, anger, surprise, fear, disgust, neutral.
On top of that we have a special model to detect drowsiness of students with good accuracy. This allows the student to know when they should take a break and drink some water or stand up so they can freshen up. It also allows the teacher to gauge the level of attention in their class.
The teacher gets the top 3 emotions present in her class as emojis displayed on her screen to be able to quickly gauge the mood of the class. Both emotions and drowsiness detection use deep learning models.
Additionally, We have also used machine learning to allow students and instructors to be able to hand draw math equations on the whiteboard, and their doodles are converted to latex code, which can present the equation in machine-written format for enhanced clarity. (Due to time constraint, this part has not been connected to the server yet) The collaborative environment also consists of a whiteboard that allows students to work together.
Log in or sign up for Devpost to join the conversation.