Digital Photgraphy Assistant A Deep convolutional neural net trained for aesthetic evalulation Summary: Gives a predicition and confidence percentage whether a photo is deemed to be aesthetic or unaesthetic.

A deep net written using the Caffe neural net framework in python, modeled after CaffeNet architecture, with weights transferred.

Final fully connected layer trained for 2 way classification of "aesthetic" or "unaesthetic"

trained on 50000 photos from rated by the community at dpchallenge.com average score of 5+ is rated as aesthetic, lower than 5 is unaesthetic.