Student Emotions helps e-learning companies make their classes better by understanding in-class emotions of their students. As collateral, that increases LTV therefore unlocking higher CAC, which helps great schools reach even more students.
In online classes every student is pretty up close and personal to the camera, so I would think it would be hard to miss someone frowning. But in a face-to-face class, students would be farther away, able to hide behind others, etc.
@oleksandr_senyuk that is true. In fact, we did a project at an off-line university two years ago. 60 classrooms were monitored and yes, only students from 1st two rows were detectable. So we ended up writing a few articles on science directories and that stopped there. Didn't further education techniques to match.
I hope this time it will work better.
@exway Well, yes. Although that was not precisely the idea behind the app. We can track attendance, emotions, leaning towards the monitor, yawning, nodding and some other types of activities.
Report
@garintheengineer Do you think that Russian companies will be interested in the new realities of promoting remote work ? )
Report
What if the students hate writing quiz? The program will suggest not to give any quiz ?
@michael_mirzoyan That is a good one, Michael.
Actually in online schools there are dedicated specialists responsible for engagement and motivation. So it is them who decide on the content.
And with this app I hope to give them more data to derive insights from.
Report
Nice idea! How good is accuracy of the emotion tracking, by the way?
@vasiliy_korshikov That is a good question. The experiments were done with the public dataset containing 980 images the API I used presents an average true positive value up to 60% (Google does 45% for reference)
Student Emotions Detector
Student Emotions Detector
Student Emotions Detector
Student Emotions Detector
Student Emotions Detector
Student Emotions Detector
Student Emotions Detector
Student Emotions Detector