AI uses emotions, facial expressions to make business sense

Our eyes and facial expressions can reveal a lot, providing companies with rich insights on how we respond to certain brands or products in a mall or even to a specific online educational course.

Ranjan Kumar, founder and chief executive of Entropik Tech, knows that very well, and uses a branch of artificial intelligence (AI) called Emotion AI to capture these data points and use AI to provide rich insights. 

In education, for instance, fatigue is induced in 60- and 90-minute video conferencing sessions, and you lose the attention of students. "While the volume of remote learning content in the market is high, their efficacy is unspecified. Emotion AI helps understand fatigue, attention and engagement levels. We analyse facial expressions and voice of students and teachers, to offer companies a solution,” explains Kumar.

Indian e-learning unicorn startup Vedantu as one of his biggest clients. Entropik Tech implemented the emotion AI solution for Vedantu in June. Kumar says demand is growing steadily in e-learning, a sector that has seen a mass-scale shift from physical to remote operations due to the covid-19 pandemic.

According to Kumar, other key sectors that are implementing emotion AI include customer service, sales and entertainment, where Entropik currently works with companies such as Flipkart, Viacom18, CavinCare and Tata Consumer Care. No wonder, he is excited that his company "has seen a four-fold growth in revenue since 2020". He attributes much of this growth to the pandemic, which enforced remote operations across industries.

To be sure, Entropik is one of a handful of companies in India that currently offer Emotion AI. 

Indian startup EnableX.io introduced face AI and analysis last year as part of its communication platform as-a-service product. The company offers emotion intelligence as a key aspect of its face AI service, which seeks to analyse facial expressions of customers to give retailers insights on which products may resonate more -- and which don’t.

Computer vision startup Mad Street Den is another player in this segment. It offers a host of AI-based solutions for the retail industry, with a goal to help retail businesses improve their sales with targeted products. Its product, called Vue.ai, studies user behaviour by tracking their usage patterns on shopping sites, and maps the same to a pool of product data to improve recommended purchases -- thereby increasing sales.

On a global scale, emotion AI businesses have been growing in popularity as well. One of the biggest global names is Affectiva, which was born out of the Massachusetts Insititute of Technology (MIT) Media Labs in 2009. In early 2021, Affectiva was acquired by Swedish advanced driver assist systems (ADAS) company Smart Eye for $73.5 million. Alongside its existing offering of using emotion AI tools for brands to improve sales, Smart Eye seeks to use the technology in ADAS systems as well.

Other big startups offering emotion AI-based products on a global scale include the Shanghai-based Emotibot Technologies, which offers emotion-recognising robots in sectors such as healthcare and education. Its most recent funding round saw Emotibot raise $31.4 million to increase its product base.

That said, while Emotion AI isn’t exactly new, the technology is being increasingly adopted by businesses. It seeks to recognise and understand human intelligence in reaction to a product or service, and offer this data in the form of analytics to product sellers and service providers to improve their topline.

Use of emotion AI is expected to grow exponentially in the years to come. A report on the use of emotion intelligence by B2B market intelligence firm Markets and Markets states, “The increase in the number of digitalisation initiatives across developing countries has led enterprises, both public and private, to deploy emotion AI-based applications. This has led to the development of new business models that may use emotional cues for business processes.”

The report states that the emotion AI market will grow at a compound annual growth rate (CAGR) of 11.3 percent to reach $3.71 billion (about Rs 2.8 lakh crore) by 2026. The Markets and Markets report claims data protection and information security regulations in an increasing number of countries are restraining the growth of emotion AI products.

In an interview with News18 in November 2020, EnableX.io’s Gupta claimed that the emotion recognition service will not identify or track individuals across the internet -- even though a business might use such services to retain emotion data linked to a customer, for future product recommendation purposes.

EnableX.io uses its FaceAI API to analyse and measure facial expressions, to offer live sales or customer service providers with real-time analysis of customer emotions -- in order to help businesses offer a more tailored customer service.

Entropik’s Kumar says his service sources data from a pool of 60 million individuals who have expressly consented to be present in this -- through third party service providers. “We don’t process any video imagery or data, and the data that we draw is emotion as a data level -- which has no personal data or information linked to it.”

Explaining how it works, Kumar says, “A brand reaches out to us with the demographic they want to target with their product, and it is then tested with that section of people from our already consented audience. Their attention, engagement and other metrics are then captured as they sample the product or service, such as an advertisement, and data from the same audience is then sent to the brand to review.”

To be sure, while builders of emotion AI products and services seem enthusiastic about their implementations in the advertisement and entertainment sectors, not all stakeholders sound as enthusiastic. 

Kainaz Karmakar, chief creative officer of Ogilvy India, is one such cirtic. He asserts, “As a creative professional, this is scary. Creativity is a fluid process, and the best results come when you go with the gut. Research has existed for ads for long, and many companies use it. However, creative folk always take it with a pinch of salt.”