Facial recognition  has been around for a while. At first and for a long time, it was only used for auto-tagging familiar faces on social media and in photo albums. Meanwhile, security agencies and law enforcement found that it could become an invaluable tool in their ongoing battle against crime fueling its development, to the dismay of civil liberties  organisation. But it is not until very recently that the field has discovered that it could bring much more, in the form of hard advertising cash.

Beyond privacy

As the social media audience matures, so does its management of privacy. Current users, unlike its first generation, are now accurately aware that not everything they do, think, or enjoy should be food for sharing online. Partly thanks to revelation from Snowden ( 30% of adults have taken at least one step to hide information from the government since his revelation) and mostly because of targeted advertising, there is an overall restraint in what they share. And that is an issue for social media platforms and their clients, advertisers.

Facebook DeepFace is 97% accurate
Facebook DeepFace is 97% accurate

Bridging relationships

People do not auto-censure their photographs as much. Especially on sites like Instagram or Snapchat where it is pretty much the only means of communication. Excited by the possibility of generating large amount social  validation, user post photos more freely than any type of text. They also assume that they are better protected against any form of snooping. Thus, they share a lot of their off-line social interactions: where we are, what we are doing and with whom. The latter is what interests researchers. Because  friends and families are the number one buying trigger, by far: 92 percent of people trust recommendations from friends and family more than all other forms of marketing. Thus, if a platform knows who they are, they can target you better. Steps-in face recognition. No longer do the platforms need to wait for you to manually identify individuals – why would you? after all you know who they are– they can do it on their own ( Facebook’s Deepface is 97% accurate), either publically or in the background. They can even automatically identify their age. And after the repeated appearance of the same individuals, it becomes easy to know who your influencers are: You’ve been mapped.

Microsoft's facial recognition engine can easily estimate age ( and gender) of subjects in a photograph
Microsoft’s facial recognition engine can easily estimate age ( and gender) of subjects in a photograph

Between you and the world

If that wasn’t enough, add a layer of content recognition, and your existence is now also linked to objects that you seem to interact frequently with. That is information advertisers love to know.

Decrypting emotions.

But face recognition is only part of the equation to solve the mystery of users. Recently, thrives  have been made in the field of emotion analysis. By reading expressions on faces, even micro expressions invisible to the human eye, machines can accurately detect  a wide variety of human emotions. They can thus link objects, location and other human beings with the emotion they trigger in you. Knowing, for example, that a beach location has you in a happy mood 90% of the time is invaluable information for anyone wanting to sell beach vacations. Or detecting that  one friend seems to appear overwhelmingly in photos where you express peacefulness is key to selling you well, pretty much anything.

via Microsoft’s Emotion demo
via Microsoft’s Emotion demo

Besides social media, emotion analysis is already being used to monitor focus groups and starting to be implemented in stores,  gaming devices and amusement parks  to better intensify your experience. A new report from Transparency Market Research estimates that the facial analytics market is set to reach nearly $2.7 billion by 2022.

Opening the gateway

And what you like is where you will spend your money. Learning who you hang out with and what you are feeling provides a wealth of information far beyond any hashtag, caption or comments. It is unfiltered, unedited data on an individual’s life that can quickly deliver actionable patterns for marketers. Via the combination of face detection, face recognition, emotion analysis and content recognition, benign photographs are now delivering high-level data to marketers wishing to deliver more accurate messaging. Consumers do not seem entirely against it: A First Insight survey found that 55 percent said they would be open to the technology if they knew a benefit was associated with it, such as discounts.



Author: Paul Melcher

Paul Melcher is the founder of Kaptur and Managing Director of Melcher System, a consultancy for visual technology firms. He is an entrepreneur, advisor, and consultant with a rich background in visual tech, content licensing, business strategy, and technology with more than 20 years experience in developing world-renowned photo-based companies with already two successful exits.

1 Comment

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.