Skip to content

Latest commit

 

History

History
13 lines (10 loc) · 555 Bytes

README.md

File metadata and controls

13 lines (10 loc) · 555 Bytes

Codes of NIMA:
An Image Aesthetics Assessment Algorithm in Pytorch

paper: NIMA: Neural Image Assessment by Hossein Talebi and Peyman Milanfar.

The datatset in the ./images is AVA (Aesthetic Visual Analysis) dataset including 250,000 images above with the aesthetic scores range from 1 to 10 for each image and some other labels such as styles and tags.

Requirements

  • Python3
  • Pytorch 0.4
  • numpy
  • scipy
  • torchvision