APSIPA Transactions on Signal and Information Processing > Vol 10 > Issue 1

Two-stage pyramidal convolutional neural networks for image colorization

Yu-Jen Wei, National Taipei University of Technology, R.O.C, Tsu-Tsai Wei, National Taipei University of Technology, R.O.C, Tien-Ying Kuo, National Taipei University of Technology, R.O.C, tykuo@ntut.edu.tw , Po-Chyi Su, National Central University, R.O.C
 
Suggested Citation
Yu-Jen Wei, Tsu-Tsai Wei, Tien-Ying Kuo and Po-Chyi Su (2021), "Two-stage pyramidal convolutional neural networks for image colorization", APSIPA Transactions on Signal and Information Processing: Vol. 10: No. 1, e15. http://dx.doi.org/10.1017/ATSIP.2021.13

Publication Date: 08 Oct 2021
© 2021 Yu-Jen Wei, Tsu-Tsai Wei, Tien-Ying Kuo and Po-Chyi Su
 
Subjects
 
Keywords
Image colorizationConvolutional neural networkImage pyramid
 

Share

Open Access

This is published under the terms of the Creative Commons Attribution licence.

Downloaded: 343 times

In this article:
I. INTRODUCTION 
II. PROPOSED METHOD 
III. EXPERIMENTAL RESULTS 
IV. CONCLUSION 

Abstract

The development of colorization algorithms through deep learning has become the current research trend. These algorithms colorize grayscale images automatically and quickly, but the colors produced are usually subdued and have low saturation. This research addresses this issue of existing algorithms by presenting a two-stage convolutional neural network (CNN) structure with the first and second stages being a chroma map generation network and a refinement network, respectively. To begin, we convert the color space of an image from RGB to HSV to predict its low-resolution chroma components and therefore reduce the computational complexity. Following that, the first-stage output is zoomed in and its detail is enhanced with a pyramidal CNN, resulting in a colorized image. Experiments show that, while using fewer parameters, our methodology produces results with more realistic color and higher saturation than existing methods.

DOI:10.1017/ATSIP.2021.13

Companion

APSIPA Transactions on Signal and Information Processing Deep Neural Networks: Representation, Interpretation, and Applications: Articles Overview
See the other articles that are part of this special issue.