Using a grayscale photograph as input, this model hallucinates a possible color version of the photograph. Used the the Caffe framework and LAB color space to implement color mapping algorithms on image inputs to colorize the image. For more details visit this article and video.
- colorization_deploy_v2.prototxt
https://github.com/richzhang/colorization/tree/caffe/colorization/models
- pts_in_hull.npy
https://github.com/richzhang/colorization/blob/caffe/colorization/resources/pts_in_hull.npy
- colorization_release_v2.caffemodel
https://www.dropbox.com/s/dx0qvhhp5hbcx7z/colorization_release_v2.caffemodel?dl=1
git clone https://github.com/mariyakhannn/imagecolorizer.git
cd imagecolorizer
To run the image colorization script, use the following command:
python main.py --image <path_to_image>
*Replace <path_to_image> with desired image path.