Earlier this year, this meme, “Puppy or Bagel”, firstly circulated via Reddit and then developed numerous variations, created a humorous viral incident in the social media sphere, but once more, signified an opportunity for the field of computer vision and image recognition (and especially in relation to advertising and marketing).
The importance of visual imagery within the digital&social media world has been tremendously highlighted over the last two years, but I believe that the last quarter of 2016 will introduce a solid direction towards more sophisticated practices of image intelligence.
Image intelligence takes a further deep dive into image analysis and focuses more on creating meaning out of the image(s) observed, as well as on the ability to create predictive models that can be later used with other data information in order to enhance business development, reinforce branding, handle crises, ride and respond to emerging trends. Deep learning is key here: a human eye can easily distinguish a puppy from a bagel even after having observed it for a long time, but for a computer software this is a complicated (and at cases, delicate) procedure. Moreover, the ability of recognising images correctly is of critical significance for future data classification. An artificial neural networks need intense training in order to ‘understand’ the complexities of image data, the differences between similar visuals, the correspondence of specific objects to specific contexts and so on.
Many global brands are now starting to realise the importance that lies behind the value of images, especially when image sharing remains so strong and drives technological advancement: According to Mary Meeker’s 2016 Internet Trends report, people share over 3 billion images every day on a range of social media platforms (mostly Facebook owned, to be precise). Observing this staggering social media trend, the research firm Markets and Markets has predicted that the image recognition field will reach almost $30 billion by 2020. According to an insightful report from Susan Etlinger at Altimeter Group, that provides independent research and strategy consulting on disruptive technology trends, rise of visual media can offer what text alone cannot. Altimeter states that in a range of interviews they did with technology providers, approximately 80% of images that included brand logos did not specifically mention the brand with any relative text.
For advertising and marketing this is a golden opportunity: Imagine how many images of all these billions that people share daily contain brands that users do not even bother to identify (and in most cases, they should not). Imagine how brand impact and influence could be measured if artificial neural networks were able to recognise every tiny detail in a massive image dataset (and what a huge development for Search). Imagine how visual brand conversations can be enhanced, taking advantage of a new communicational era that corresponds through visuals instead of text. Generation Z is now overtaking Millennials, finding new ways of visual communication, speaking a visual language that is not yet decoded. However, imagine how facial recognition could help towards emotion detection and analysis. Or how brand storytelling, brand affinity, competitive benchmarking, influencer marketing, content marketing and so on, could be developed. This whole new world cannot be missed.
Of course, challenges are in the agenda and the first one is what troubles advertising and marketing field over the last few years: the transition from traditional practices to a highly technological and innovative era. The advertising market is already struggling to swallow a massive shift towards technologically advanced practices (data science, big data analytics, social analytics, programmatic advertising, digital advertising, and so on) with many companies stating how the lack of technological skills in their workforce, complicates their business. Moreover, data giants like Facebook, Google and Amazon are constructing a new reality that does not include the old advertising world; a world that lately seems to be standing puzzled in front of seismic changes.
Thus, introducing terms like ‘artificial intelligence’, ‘deep learning’, ‘neural networks’, ‘virtual reality’, ‘augmented reality’, ‘image analytics’ might sound scary, however this trend is here to stay. Those organisations realising that the only way forward is to ride it, will be shaping the developments of the (visual) future.
Image source: http://www.boredpanda.com/dog-food-comparison-bagel-muffin-lookalike-teenybiscuit-karen-zack/