Virality Detection API gives a score to images based on their potential to go viral on the Internet. This API has been bulit by training a super deep neural network it on a huge corpus of image and their scores crawled from the open web. The score that you get as output is the virality score of the input photo out of maximum score of 100. Our in house experiments suggest that accuracy of this algorithm to predict image virality is as high as 85%.
We are soon coming up with a paper for the methods we use here which will be available as open access publication.