Artificial Intelligence has taken over recently but now it has entered the education field. Online classes can seem tedious for the teacher and students; technology has taken steps to improve this process. What if there is a system that can monitor students based on how long they take to answer questions and reports their performance? A Hong Kong company has made this happen. Find out how.
How Can AI Read Emotions?
Communication was easier before the pandemic; the teachers could gauge what the students are up to. With online classes, it has become difficult for the teachers to see the students and communicate effortlessly. Ka Tim Chu, a teacher, and vice-principal of Hong Kong’s True Light College says that technology is helping him read the room. AI has come up with a solution to monitor the student’s emotions in online classes.
4 Little Trees is an AI-powered learning platform created by a Hong Kong-based start-up. The founder, Viola Lam found a unique solution for better communication in virtual classes by making it better and real. Emotional recognition AI has been a concern but the start-up is trying to bridge the gap.
Lam founded 4 Little Trees in 2017 with a funding of $5 million; her initial thought process was to give teachers an ‘intervention.’ There has been a rise in the number of schools that are using 4 Little Trees in Hong Kong- it jumped from 34 to 83 last year. Lam found out that the technology was useful to teachers in the pandemic as it can allow them to figure out the student’s emotions while learning. The prices of the start-up range from $10 to $49 per student and per course.
Interesting Way Of Learning With AI
Studying hasn’t stopped because of the pandemic, students continue on with classes and tests. The AI is designed in a way in that the muscle points on their faces through the camera on the computer or device can identify their emotions such as happiness, sadness, anger, or fear. What is fascinating is that the system can monitor a student’s complete performance from them answering questions to generating reports on their strengths and weaknesses.
The unique feature of this technology is that it can make personalised game-style tests for each student which makes it fun for them. Lam stated to CNN that students perform 10% better in their exams if they learn through 4 Little Trees.
Technology has played a huge role in reducing the distance between teachers and students. It has made a teacher’s life easier with personalized tests and assignments; it lessens the burden upon them. The expression reading AI is significant as every student’s emotions are monitored in a huge class. This would not only help in online classes but when the pandemic ends, it could be useful for schools and colleges as a long-term benefit.
Will Privacy Be Breached With AI?
The personal attention on every student can make things easier for the teachers. There are some concerns in terms of privacy. In China, AI analyses biometric data for surveillance purposes which raised questions on how would this work. Lam was quick to dismiss the controversy, she said that 4 Little Trees records facial muscle data and interprets emotional expressions but it does not video student’s faces.
The sole purpose of the AI is to track the movement of muscles on the student’s faces to read their emotions. If the corners of the mouth are raised, we can detect that the student is happy. Assessing the emotions of the student to understand and communicate better in online classes is the main reason.
The director of the center of AI Research at Hong Kong University of Science and Technology, Pascal Fung believes that transparency is important in maintaining student’s privacy. She reiterates that the developers should receive consent from the parents to collect the student’s data and mention what the data is going to be used for.
Racial Bias Is Predominant In AI
Another concerning issue was racial bias in AI. Research has been done which shows that technology finds it difficult at times to identify emotions of dark-skinned faces because the algorithm is founded in a way of human bias. It has been learned only to identify mostly White faces which makes it a racial concern
Lam said in an interview clearing the air that the AI is well-trained to identify facial data of any student of any demographic. As it is predominantly based in Hong Kong, it could be a challenge to detect ethnically-mixed communities. The AI recognition works with 85% accuracy in Hong Kong which is good as it can identify primary emotions such as happiness or sadness easily. There is still room for improvement to identify complex emotions such as irritation or anxiety.
Fung mentioned in an interview that they have been hoping for a 50-60% accuracy to identify complex emotions. They plan that machines would be better than average humans to read facial expressions and with technology today, it is already happening.
They are working on improving the AI to develop this technology not only for schools but businesses and increase its engagement in online meetings and webinars. Lam believes that AI can help facilitate a better interaction when compared to human communication.
This turned out to be an interesting way for bridging the gap between students and teachers during the pandemic. With the rise of technology, communication has become easier and this venture could help build a better learning experience for students.