A professor in China is using facial recognition to gauge student interest

A computer science professor at China’s Sichuan University is using facial and expression recognition technology to gauge the interest levels of students in his classes.

Professor Wei Xiaoyong’s self-developed “face reader” is used to tell how stimulated his students are — and to utilize this data to improve his teaching.

“Five years ago, when I came to this university to teach, I came up with the idea of using facial recognition for attendance checks,” Wei told Digital Trends. “Over time, I realized there was other information we could gather from this data. For example, it is possible to figure out social networks in the classroom by looking at which students sat next to each other.”

The research grew from here, and Wei decided to move on to monitoring not just faces, but emotions as well.

“What we’re doing now is to use an expression recognition system to figure out students’ behavior in the classroom,” he continued. “He hope this can be used to carry out analysis, for example telling whether we’re teaching in a way that is optimal for them. It might be possible to spot patterns in behavior so we know what to do next in a certain scenario — like knowing when to tell a joke to attract their attention.”

Wei noted that students in his class seemed more intrigued than unsettled by the technology — and added that he is renowned for using unusual methods in his teaching. “Most of the students are coming for that, so they are prepared,” he said. “We also offer them three weeks to get used to new classes, so if they don’t like the system they can drop the class. After that three week period, we ask them to sign a consent form for video.”

At present, he described the expression recognition system as being at a relatively primitive stage. “We only track two expressions: happy and neutral,” he noted. “But it’s not a technical problem for us to add new expressions and emotions. That’s something we’d like to do over time.”

The system is also currently trained on offline data, which means that Wei is not able to get real-time feedback about his classes while they are happening. Instead, insights can be gathered between classes to gradually improve his students’ learning experience.

Personalized education revolution or dystopian future? You tell us.