The University of San Diego wants to maximize their likely future interactions between students and teachers robotic. The goal: to have a robot that pace in its current function of facial motion of its audience
Facial expressions should be able to control the behavior of a robot. Jacob Whitehill, a researcher at the UC San Diego School of Engineering, works in the remote control of a video sequence showing the course depending on the degree of understanding of students. The project in the idea that future teachers robotic systems could be developed. The long term goal is to make robots better teachers. The project is based on technology of facial recognition. Thus, Jacob Whitehill has shown that the information given by our facial expressions while watching videos help you understand whether or not the explanation which is filmed.
Eye blinks of significant
He explains: “If I am a student with me in front of a teacher, and I’m lost at the demonstration, my face will show him.” “It will then stop and say” you look confused, I’ll slow down my speed. ” Its goal is to have the same type of interaction with a robot teacher. To do this, first we need to collect the different facial movements that students can do. The researcher looked a group of students taking courses in German grammar, and filmed all the reactions of students during these lessons. It appears that everyone has a variety of expressions. However one can observe a common trend.
Optimizing the relationship pupil / teacher
Most students clignent eyes at least portions of the course the more complex compared to the most easy to follow and understand. The next step will be to determine what facial movements one person performing naturally in a situation where it is exposed to a complex explanation. “My goal is to optimize the interaction student / teacher,” says Jacob Whitehill. “I want to succeed in capturing the expressions expressed by students who stumble on an idea, and those teachers who want to enrich their demonstration by example”. To ultimately develop robots that punctuate their courses based on feedback from their interlocutors.