Emotions of children can now ‘be read’

Budrigannews.com – Ka Tim Chu, a teacher and vice principal at Hong Kong’s True Light College, looked at the faces of his students before the pandemic to see how they were responding to the work. Chu is now able to read the room thanks to technology, as the majority of his lessons are now online. His students’ emotions are monitored by an AI-powered learning platform as they study at home.

Find Solution AI, a startup based in Hong Kong, developed the software, 4 Little Trees. Viola Lam, the company’s founder, claims that emotion recognition AI can make the virtual classroom as good as or better than the real thing, despite concerns about its use in educational settings and other settings.

As part of the school curriculum, students use the platform to work on tests and homework. The AI identifies a variety of emotions, including joy, sadness, anger, surprise, and fear, as they study by using the camera on their computer or tablet to measure the muscle points on their faces.

Additionally, the system tracks how long it takes students to answer questions; records their performance history and marks; produces evaluations of their strengths, weaknesses, and levels of motivation; and anticipates their grades. The program can be tailored to each student, focusing on knowledge gaps and providing fun learning games and tests. According to Lam, students who have learned through the use of 4 Little Trees perform better on exams by 10%.

Lam, who used to be a teacher, remembers discovering that some students were struggling only after they received their exam results, at which point “it’s too late.”

With $5 million in funding, she launched 4 Little Trees in 2017 to provide teachers with an opportunity for “earlier intervention.” Over the course of the past year, there have been an additional 83 schools in Hong Kong that use 4 Little Trees. Prices for each course range from $10 to $49 per student.

According to Lam, the technology has been particularly helpful to educators during the pandemic because it has allowed them to remotely observe the emotions of their students as they learn.

Chu believes that the technology’s advantages will outweigh the pandemic because it lessens his administrative burden by allowing him to create and mark personalized assignments and tests. In addition, in contrast to teachers, the expression-reading AI is able to pay close attention to every student’s feelings, even in a large class.

However, technology that monitors children’s faces raises privacy concerns.

AI that analyzes biometric data for the purpose of surveillance in schools and other locations has sparked controversy in China.

According to Lam, although 4 Little Trees does not take videos of students’ faces, it does record facial muscle data, which is how the AI interprets emotional expressions.

“Transparency” is essential for safeguarding students’ privacy, according to Pascale Fung, director of the Center for AI Research at Hong Kong University of Science and Technology. She asserts that in order for developers to collect student data, they must first obtain parental consent and “explain where the data is going to go.”

AI also faces a serious problem with racial bias. According to research, some emotional analysis technology has difficulty recognizing the emotions of faces with darker skin tones. This is due, in part, to the fact that the algorithm is influenced by human bias and learns to recognize emotions from predominantly White faces.

Lam claims that she uses facial data that matches the students’ demographics to train the AI. Although it has been effective so far in Hong Kong’s predominantly Chinese society, she is aware that the software may face greater difficulties in communities with a greater ethnic diversity.

Emotional expression can vary by culture and ethnicity, according to experts.

Lam claims that Find Solution AI’s emotion recognition is effective in Hong Kong with an accuracy of 85%. Up to 90% of the time, according to Fung, algorithms with “very good settings” can correctly identify primary emotions like happiness and sadness.

However, it can be more challenging to read more complex emotions like annoyance, enthusiasm, or anxiety.

Fung states, “We can hope for 60% [or] 70% accuracy,” noting that the majority of people are unable to identify complex emotions with greater accuracy. She asserts, “Humans are not good at reading facial expressions.” We want to teach machines to be better than the average person.

More Artificial intelligence will control the scooter

Lam hopes to develop applications for schools and businesses to better understand the needs of participants and increase engagement in online meetings and webinars as the AI gets better.

More Artificial intelligence warn you before water leaks

She asserts that AI “can help to facilitate a better interaction” when it comes to human communication.