Skip to content

Latest commit

 

History

History
28 lines (17 loc) · 792 Bytes

README.md

File metadata and controls

28 lines (17 loc) · 792 Bytes

Automatic Image Colorization

This is a Deep Learning academic project, throughout this project we explored state of the art deep learning techniques for automatic colorization of gray scale images. For further details check this report.

Methods used

  • DDColor
  • Conditional Generative Adversial Networks (cGANS)

Dependancies

  • PyTorch
  • tqdm
  • pandas
  • dotenv
  • skimage
  • PIL
  • numpy

Dataset

The code expects the COCO-2017 dataset stored in the coco-2017-dataset directory . You can set the path for this and other directories in a .env file.

.env File

Follow the structure of .env.template