Dubai students create AI face-reading tool that tells teachers if classes are boring

The technology is designed to improve distance learning by measuring engagement and delivering feedback

Two students at Murdoch University Dubai have developed an artificial intelligence-based tool that scans pupils’ faces during classes to gauge their level of understanding.

The application aims to help improve the quality of online education by acting as a teaching tool and providing live feedback on whether the lessons are engaging.

It analyses facial expressions such as eyebrow raising, eyelid tightening and mouth dimpling to measure whether students are interested in the class.

Quote
We were tired of how classes were being held and no one was paying attention
Eiman Ansari

Eiman Ansari, 20, is in the third year of a degree in information technology and cyber forensics.

She said she was frustrated by online classes because teachers sometimes struggled to gauge students’ reactions.

“We were tired of how classes were being held and no one was paying attention,” said Ms Ansari, who is Indian.

“When classes were held, our cameras were off and the interactive classroom experience had disappeared.

“The teachers were not able to see the students’ reactions so they did not know if we were understanding things correctly.”

Ms Ansari developed the application with Abubakr Sajith, 21, a third-year undergraduate in business information systems and finance.

The friends were supported in their research by Joseph Stevens, the university’s head of information technology.

Measuring fear, disgust and surprise

Their application, the Emotion-Affective Domain Mapping System, films the classroom and analyses facial expressions for emotions such as contempt, anger, fear, happiness, disgust and surprise.

A graph is produced to illustrate students’ level of engagement in the class.

“We are using artificial intelligence to analyse facial expressions and the data given from that is mapped to levels of understanding,” Ms Ansari said.

“The aim is for students to have education tailored towards them, while for teachers this is a helper tool.”

The duo developed the idea in November 2020. Their research has since been published in The International Journal of Inspired Education, Science and Technology.

Their current paper proposes the system. Ms Ansari and Mr Sajith are working on a second paper to measure its effectiveness.

Mr Sajith said that while online classes would not last forever, the technology could have a place in face-to-face classes as well.

“We would ideally see this in a class where teachers have a dashboard on the side along their table and they can see a live graph showing how their pupils are reacting,” said Mr Sajith.

“With each student, you see the method they react to, and which one suits them best.”

At present, teachers will be able to gain feedback after the class. Having reviewed their students’ reactions, they can modify their class content and delivery.

In the future, teachers would be able to receive live feedback during the class.

Mr Sajith said that although the technology was in its nascent stage, the team hoped to one day license it to universities and schools.

Dr James Trotter, Murdoch University Dubai’s dean and academic president, acknowledged that measuring students’ comprehension during online lessons was hard.

“It is also estimated that the problem of poor integration of technology within the education systems is set to cost us roughly 0.3 to 0.9 years’ worth of education per year,” he said.

“The proposed EADMS solution, by two Murdoch University Dubai students and their faculty mentor, is a novel way to improve comprehension assessment and subsequently improve course content and delivery.”

Updated: August 11th 2021, 9:58 AM
EDITOR'S PICKS
NEWSLETTERS