Epitome for Automatic Image Colorization

TL;DR

This paper introduces an automatic image colorization method using epitome, a generative model that summarizes image appearance and shape, enabling effective colorization without human input.

Contribution

The novel use of epitome as a condensed model for automatic colorization eliminates the need for user interaction and improves colorization quality.

Findings

Outperforms previous methods in colorization quality

Automatically colorizes grayscale images without user input

Effective summary of color information via epitome

Abstract

Image colorization adds color to grayscale images. It not only increases the visual appeal of grayscale images, but also enriches the information contained in scientific images that lack color information. Most existing methods of colorization require laborious user interaction for scribbles or image segmentation. To eliminate the need for human labor, we develop an automatic image colorization method using epitome. Built upon a generative graphical model, epitome is a condensed image appearance and shape model which also proves to be an effective summary of color information for the colorization task. We train the epitome from the reference images and perform inference in the epitome to colorize grayscale images, rendering better colorization results than previous method in our experiments.