Single Image Colorization Via Modified Cyclegan

In this paper, we focus on automatically colorizing single grayscale image without manual interventions. Most of existing methods tried to accurately restore unknown ground-truth colors and require paired training data for model optimization. However, the ideal restoration objective and strict training constraints limited their performance. Inspired by CycleGAN, we formulate the process of colorization as image-to-image translation and propose an effective color-CycleGAN solution. High-level semantic identity loss and low-level color loss are additionally suggested for model optimization. Our method allows using unpaired images for training and direct prediction in rgb color space, which makes training data collection much easier and more general. We train our model on randomly selected PASCAL VOC 2007 images. All ablation study on loss function and comparisons with state-of-the-art methods are performed on grayscale SUN data. The experiment results show that our improvements on training loss could achieve better content consistence and generate better reasonable colors with less artifacts. Moreover, due to the bidirectional nature of our model, our proposed method provides a by-product that gives an excellent alternative way on color image graying.

1. INTRODUCTION

Automatic colorization is a technology that can automatically colorize gray image without manual intervention [1]. Since it has great practical applications, such as colorizing old black and white photos, artistic creation etc, it has always been an active topic in image processing research field. However, as many colors share similar gray value, recovering original col-or is an ill-posed problem. Therefore, automatic colorization is also a very challenge task.