Deep Learning Photo Colorization and color space learning. Exploring research literature and implementing prototypes using Pytorch
Basen on LAB color space.
Amount of light on three zones (dark, mid, light) in RGB channels simultaneously. Working with the prediction of color depends on light/color zone.
Results: simple convNN (7 layers)
The second form of data augmentation consists of altering the intensities of the RGB channels in training images. Specifically, we perform PCA on the set of RGB pixel values throughout the ImageNet training set (ImageNet Classification with Deep Convolutional Neural Networks paper)