Critique
Not Good Enough: Neural Network Evaluates Iconic Photographs
Bird In Flight uploaded iconic photographs into the Everypixel Aesthetics application, which is designed to evaluate the awesomeness of the image. We got some unexpected results.
Everypixel presented their development — a neural network called Aesthetics, which predicts the ‘chance that the image is awesome’ in percent. The neural network was taught to predict accurately with the help of 957,000 photographs of different quality. Bird In Flight decided to check the accuracy of Everypixel Aesthetics by feeding it the classics of photography.
Cindy Sherman
Helmut Newton
William Eggleston
Pierre et Gilles
Mike Wells
Sam Shere
Martin Parr
Wolfgang Tillmans
Charlie Cole
Gregory Crewdson
Rineke Dijkstra
Boris Mikhailov
Nan Goldin
Andreas Gursky
Roger Ballen
Alexander Rodchenko
Nick Ut
Kevin Carter
Terry Richardson
Joe Rosenthal