review site logo
March 16,  2017

Everypixel Measures Visual Appeal in Stock Images through Neural Network

By Topimagessites

 Everypixel, a new stock image search engine, is beta testing its unique algorithm which is trained to check 'awesomeness' of stock images.

This 'awesomeness' is defined according to the visual appeal of an image for the neural network of a human.

You can upload an image directly or provide an URL, and it works not only on stock images but any image. The algorithm is designed to give you a percentage of how good the image is.

It has been observed that run-of-the-mill selfies score lower than images of landscapes, diversity, and works of art.

The platform also has an additional feature of suggesting tags. It suggests accurate tags for images after analyzing them.

The algorithm is developed by a team of designers, editors, and experienced stock photographers at Everypixel.


To read the full article, please go here.