As two major breakthroughs in the world of image recognition software begin to pique the interest of online marketers, artificial intelligence starts to prove its use in the world of advertising ever more.
A team of scientists at Google have developed artificial intelligence software that can describe the contents of photographs and accurately analyse sentiment, providing descriptions via natural language processing.
The new self-learning machine uses two neural networks, utilising a combination of image recognition software and natural language processing to describe exactly what’s happening in a photograph by recognising the subjects and the objects they may be interacting with.
Originally designed to help improve the quality of search engine results, Google has stated that the new AI software could be used to describe pictures to blind people, but the possibilities of its use in the online marketing world could prove extremely useful in the near future.
This picture was automatically captioned: “Two pizzas sitting on top of a stove top oven”
Taking this image mining a step further; Ditto Labs have also created AI software that can recognise brands, products and the location of a picture, whether in a snowy mountain or a bar for example, and identify sentiment by scanning the faces of the person or persons in the photograph giving advertisers an insight into the type of people they should be marketing to or information on how their products are used.
For example, if a photo is taken of you whilst you’re snowboarding and you’ve got a coke can in your hand, the new AI software will recognise that you’re a fan of outdoor activities and drinking coca cola. Ditto will then relay that information to advertisers, allowing them to target you with specific ads based on your preferences and the sentiment of the photo, whether positive or negative.
Photo: Wall street journal
These breakthroughs in sentiment analysis and image mining pose new and exciting opportunities for marketers and brands, allowing for an even greater depth of sentiment analysis and providing brands with the information they need to target their demographic more effectively.
To be continued...