/Image_Colorization_Using_CNN

Primary LanguagePython

Image_Colorization_Using_CNN

Objective

Given a Grey-Scale Image, predict its colored version

Approach

We use a U-Net kind of architecture where we feed Grey-Scale Image as Input, and generate AB Channel as output. Final Predicted image is concatenation of output AB channels and input grey-scale image.

Instructions to run

  1. Link to dataset : https://www.kaggle.com/namanj27/images-places
  2. Save the dataset (‘images’ folder) to Images_Places/ directory
  3. Run the Code.py file