Affectiva, a Smart Eye company, today announced the latest version release of its category-defining Emotion AI product. This update to its media analytics offering makes several new features available to its market research customers.
Marketing Technology News: Martech Interview with Li Jun, Founder of Ontology
“It represents our continued commitment to the media analytics industry through the iterative improvement of our technology, not to mention demonstrates the groundbreaking work our team has achieved in helping our customers further understand complex and nuanced emotions and cognitive states with technology.”
Affectiva’s Emotion AI technology is used by 70 percent of the world’s largest advertisers and 28 percent of Fortune Global 500 companies to understand consumers’ and customers’ emotional reactions to content and experiences, maximizing brand ROI. With the help of Emotion AI, clients can test the unbiased and unfiltered emotional responses that consumers have with brand content such as video ad content and longer TV programming. The device-agnostic system works across mobile, tablet, desktop and physical environments, and works with optical sensors, standard webcams, near infrared and RGB cameras.
Affectiva has continuously invested in its media analytics and Emotion AI capabilities both leading up to and after being acquired by Smart Eye in June 2021. The update makes several notable improvements to its existing Emotion AI offering:
- Affectiva’s previously best in class face tracker has furthered its accuracy even more to understand the precise placement of faces on-screen, regardless of lighting conditions and head angles. Drawing on work done for automotive safety in very challenging camera and lighting conditions, this update greatly increases the usable video rate, increasing analytical robustness and cutting costs for clients.
- More substantive, enhanced facial emotion metrics that pick up on consumer and audience emotional states more accurately (such as anger and fear). This allows for a more meaningful and precise measurement of unfiltered and unbiased emotional and cognitive responses to content, all captured with opt in and consent.
- The introduction of sentimentality and confusion metrics. Measuring these two states has been largely difficult up to this point, as very little literature exists regarding the facial signals of these two mental states and underlying facial expression science. With understanding these two states a common request among clients, Affectiva had to break new ground in identifying relevant expressions that map to sentimentality and confusion. As a result, no other facial expression analysis system identifies these two mental states. These capabilities will be particularly relevant for Affectiva entertainment research clients.
Marketing Technology News: VisioNFT Token Bringing Virtual NFT Galleries To The BSC
The new emotion AI generates more unique insights that can then be practically applied to improve brand experiences and communications. This builds more positive and meaningful connections with customers, and enables optimization and action. It also directly impacts the bottom line as clients can now generate more usable data from the same sample number of responses; for example, the face tracker improvements can be directly applied to the increasing use of mobile phones in data capture, where data quality has historically struggled due to low lighting conditions and challenging head angle positions.
This SDK update also draws upon a massively diverse data set for algorithm training. Today, Affectiva has the largest emotional database in the world, with over 11.5 million consumer responses to more than 60,000 ads in 90 countries. It also reflects the most global representation of diversity in a media analytics data set, mitigating algorithmic bias and increasing cross-cultural sensitivity.
“I am especially proud of how Affectiva has been able to draw upon the diversity of our training set,” said Dr. Rana el Kaliouby, former Co-Founder and CEO of Affectiva, now Deputy CEO of Smart Eye. “For example, the majority of our data set is not Caucasian today, which allows us to minimize bias more effectively. This aligns with our very intentional mission to mitigate algorithmic bias, and to further the cause of prioritizing ethics in AI.”
Marketing Technology News: Diebold Nixdorf Names Microsoft Corp. Executive Bill Borden as Company’s Newest Board Member