Now, Facebook's AI engine will tell you what to wear
Engineers around the world are using Artificial Intelligence (AI) for cancer detection, self-driving vehicles, and many other innovative solutions aimed at making our lives easier and more convenient.
And, now social networking giant Facebook is using its neural engine for something completely weird and unexpected - to help you dress up in a fashionable way.
Here's all you need to know about it.
Fashion++ AI to make your outfit fashionable
The way one dresses up can be subjective but Facebook wants to offer help in that area.
The company's AI team has developed a system, called Fashion++, that offers recommendations to make an outfit more attractive.
However, unlike other systems in this arena, it will only suggest easy changes to an existing outfit, instead of entirely new clothes or similar ones.
What kind of alterations the AI suggests
The deep image-generation neural network of the system recognizes an outfit and offers minimalistic alterations, changes that make you look better and are more practical than buying a new outfit.
For instance, it could tell you to add/remove/swap a certain item or make changes to how it's worn, like when a shirt should be tucked in or sleeves should be rolled up.
It has been trained on thousands of images
Fashion++ offers outfit recommendations with its discriminative fashionability classifier, which has been trained on a dataset of thousands of images that have been deemed to be stylish and fashionable.
It uses this dataset as the ground example for adjusting the looks, and according to human evaluations, the suggestions from the system are not only fashionable but also easy to implement.
This could help designer create new looks
Facebook's AI could lead to the development of apps capable of suggesting how to make an existing outfit more fashionable and attractive for different occasions.
More importantly, the tech could offer designers a way to create and share completely new looks, styles that are yet to be discovered.
However, it remains to be seen how long this tech would take to reach that point.