With the rising popularity of digital personal assistants, natural language processing (NLP) technology has been getting a lot of AI attention. However, there is another natural sense that we need to consider giving to computers; the sense of sight. Computer vision is a red-hot topic in numerous recent applications of Machine Learning. The most famous examples would be Tesla’s autopilot and Apple’s face ID released with the iPhone X in late 2017.

Bringing our focus back to marketing – we should ask ourselves: is there anything in this industry that can benefit from computer vision? Indeed there is. Imagine when you’re browsing those cute kitten images on Pinterest. The image itself would be a good place to advertise cat food (this is called in-image ads or contextual ads). With computer vision algorithms that can distinguish not only just a cat but also “Siamese cat”, advertisers now have the ability to target specific image contexts to deploy their ads.

Sure, all of this covers the most advanced digital ad deployment methods, but what about traditional outdoor ad posters? Unlike digital marketing, outdoor posters are becoming a little out-of-date since there is no way to track “users” and “number of views” like their digital counterparts. A viewer who glances at a poster and a viewer who stares at a poster for five-to-ten seconds are totally different. The solution could be something being dubbed now as an Artificially Intelligent Poster, invented to specifically tackle this problem. By tracking the viewer’s facial expressions using computer vision, marketers are able to assess and potentially even deliver ads based on reactions from people through simply looking at a poster!

In the future, advertisements on desktop or mobile may also gather details about attention span and emotional response through facial recognition, and potentially even start to take into account more intricate data-sources like breathing patterns, body temperature (through hands), and much more.

To conclude – consider the following thoughts: How do we determine what ads should be shown next? What details should the advertisers themselves change? How does one ‘opt-out’ of this sort of advertising? All of these developments, while exciting, will undoubtedly continue to create even more hurdles for protecting user privacy in the not so distant future.