By Susana Galeano
LATAM Content Coordinator
ORA cat, a car, a busy street, the Eiffel Tower or La Giralda: surely you, like millions of people on the planet, are able to recreate a mental image of them when you read these words, and if we showed you a photograph you would be able to differentiate each visual object that makes it up in a single glance. Something that seems so simple and normal to us requires, among other things, millions of neural connections and prior learning to be carried out, something that SEER, Facebook’s artificial intelligence designed to “see” any image, also wants to achieve.
Facebook has created a new artificial intelligence system that it hopes will be much more efficient for companies to perform a range of computer vision tasks, from facial recognition to some functions greece business email database necessary for the operation of autonomous cars, according to a series of publications that the company has published on its blog, and according to Facebook.
1 billion Instagram images have fueled SEER
SEER has been trained on over 1 billion public Instagram images , and according to the company, it has managed to outperform existing AI models in an object recognition test with a score of 84.2%.
Babies, mountains, cats, lunches, families… Instagram has become one of the largest image databases in the world over the last decade, and of course Facebook has known how to use this great treasure to feed SEER and learn what is behind each of these images.