Generating colorized images based on interactive color edges using colorized neural networks

Abstract

Embodiments of the present disclosure relate to generating colorized images based on interactive color edges using colorized neural networks. Methods, systems, and non-transitory computer-readable storage media are disclosed for converting a grayscale digital image into a colorized digital image using an edge prediction neural network and an edge guided colorization neural network. In one or more embodiments, the disclosed system applies a color edge prediction neural network to a grayscale image to generate a color edge map indicative of predicted chroma edges. The disclosed system may present a color edge map to a user via a colorized graphical user interface, and receive user color point and color edge modifications. The disclosed system may apply a second neural network, an edge-guided colorization neural network to the color edge map or modified edge map, user color point, and grayscale image to generate an edge-constrained colorized digital image.

Description

technical field [0001] Embodiments of the present disclosure relate to the field of digital images, and more particularly to generating digital images that reflect selected colors. Background technique [0002] Significant improvements have been made in recent years in the hardware and software platforms used to edit and generate digital images. For example, developers have created techniques for modifying or improving digital images, including photographs and illustrations. To illustrate, conventional image colorizing systems can automatically add color to complex black-and-white or grayscale images based on user selection of a particular fill color. For example, a client device can select a color that corresponds to a particular area of ​​a grayscale image, and the image colorization system can generate a digital image that reflects the selected color. [0003] Although conventional image colorization systems can add color to digital images, they suffer from several tech...

AI Technical Summary

Problems solved by technology

Consequently, conventional image colorization systems are often inflexible and fail to provide fluid options for modifying the colorized digital image and shaded regions

[0005] Furthermore, conventional image colorization systems are often inefficient

In particular, conventional systems must often devote additional computing resources to correct for bleeding or other inaccuracies

Images