
Share
1st July 2015
06:03pm BST

A report provided by the Computer Vision Foundation explains: "People want to look good; business or casual, elegant or sporty, sexy but not slutty, and of course trendy, particularly so when putting their picture online.
"Our goal is to learn and predict how fashionable a person looks on a photograph and suggest subtle improvements the user could make to improve her/his appeal."
Researchers at the University of Toronto examined 144,169 photos from chictopia.com and amassed highly detailed statistics from each post, including the fashion it features and the comments and engagement it received.
The data, known as 'Fashion144K Dataset' revealed certain patterns between the characteristics of a post and the interest it generated.
The brainboxes at Toronto then crunched the data into their algorithm.
When the user inputs a photo, the tool will assess "the type of outfit and garments, the type of user, the setting/scenery of the photograph, and fashionability of the user's photograph."
Then, the feature will provide a set of feedback including outfit recommendations, plus suggestions on the location, lighting and angle of the photo.
Kendall Jenner watch out, we're coming after your record.
