Kunal Kislay is a B.Tech IIT Bombay alumnus with over a decade of experience in enterprise mobility, Internet of Things, AI, Neural networks, and Machine learning. He co-founded IWiz – a made-in-India, AI-enabled Computer vision platform helping enterprises and B2C segments. Their client relations are built across continents with AWS, Microsoft, Xerox, Godrej, Heineken, Johnson Controls, Tata, etc.
A complete shift to online modes of learning during the pandemic-induced lockdowns was overwhelming not just for students but also for the teachers. In addition to logistical and network issues, keeping the students engaged also became challenging for teachers. This was primarily because teachers couldn’t read the faces of students as they did in physical classrooms. But as they say — there’s always a solution. An artificial Intelligence-powered technology came to the rescue. The technology read the facial expressions of the students and helped teachers understand the level of their engagement.
Over the last few years, the usage of AI for facial recognition has gained prevalence in a wide range of areas, including security, in a very short span of time. As human-machine interaction becomes more evolved, facial recognition technology is being assimilated in several activities — from unlocking our phones to accessing personal documents on our devices — of our day-to-day lives.
Going a notch higher, AI is now detecting our facial expressions to ensure more safety and convenience. Today, there are AI-based software that measure muscle points on faces and identify emotions like sadness, happiness, anger, fear, disgust, contempt, and surprise.
Such software monitor the position and movements of brows, eyes, mouth, and other features of a person and then compare the facial expression to an already learned emotion. For instance, a wrinkled forehead and an upside-down smile are seen as a characteristic of a sad face by the software. In fact, the AI facial emotion recognition technology isn’t just capable of grasping how a person is feeling in the moment but also predicting their intentions and decoding their personality.
Use cases of AI for facial recognition
As mentioned earlier, this sophisticated tech is now finding its importance in almost all walks of life. In the context of security, AI is being leveraged to prevent road accidents. Top models of several cars have started to offer an AI feature that alerts the drivers of their drowsiness, reducing accidents by a significant percentage. The software reads the driver’s face to identify if the person is distracted, drowsy, yawning, or closing eyes, among others. Some of these software even send personalised alerts — like stopping for a coffee break, or changing music or the car temperature — to the driver.
In the medical sector, AI emotion recognition technology helps medical professionals understand the patients who have trouble expressing themselves through normative means.
Similarly, the technology also facilitates the recruitment process by assisting the HR person in understanding a candidate’s facial expressions. This helps employers assess the personality and mood of candidates. According to media reports, Unilever is already starting to incorporate this technology into its recruitment process.
Emotion detection from facial expressions using AI also provides a good way for market research. Companies are already using sentiment analysis to gauge consumer mood towards their product or brand in the digital world. After shopping online or at a retail store, we are often asked to pick an emoji — among angry, unhappy, satisfied, and happy faces, and others — to show the level of our satisfaction with the brand. Going forward, the AI emotion recognition technology will be a viable alternative to automatically measure consumers’ engagement and experience with the brand.
Further, video gaming companies can use the technology for their benefit. As each video game aims to evoke a particular behavior and set of emotions from the users, AI-based emotion detection software can help in understanding which emotions a user is going through in real-time as he is playing.
The technology is also being used in a big way in the area of law enforcement. In the United Kingdom, the Lincolnshire police have even invested in this AI-based emotion recognition to identify “suspicious” people. Other use cases of AI facial emotion recognition technology include — surveillance, & monitoring, advertising, and entertainment, among others.
Long way ahead
Despite the giant strides the technology has made in the last few years, AI-based emotion detection is still in its nascent stage. Sometimes these software falter in detecting gender and race. There are also concerns regarding privacy and mass surveillance around the technology. Also, some experts argue that technology works on the concept that people express emotions in a similar manner everywhere. They point out that people emote in different ways based on their culture and societies. In 2019, a study by the Association for Psychological Science also proved this point.
Nevertheless, almost all observers are unanimous that facial recognition powered by AI is going to be a multi-billion-dollar industry in the future. According to market research firm “Markets and Markets”, the emotion detection industry — which was valued at $19.5 billion in 2020 — is projected to grow to $37.1 billion by 2026.
Today, more and more start-ups are pushing back against the idea of a universal set of expressions and are finding opportunities to interpret a section of people with context. This process of comprehending human emotions is crucial to making AI robust. Future innovation in emotion recognition will allow machines to understand how people feel and this will truly revolutionise man-machine interaction in the times to come.