5,337
edits
SinGAN: Learning a Generative Model from a Single Natural Image: Difference between revisions
SinGAN: Learning a Generative Model from a Single Natural Image (view source)
Revision as of 17:26, 12 November 2019
, 12 November 2019→Applications
No edit summary |
|||
| Line 89: | Line 89: | ||
==Applications== | ==Applications== | ||
The following are applications they identify. | The following are applications they identify.<br> | ||
The basic idea for each of these applications is to start at an intermediate | The basic idea for each of these applications is to start your input at an intermediate GAN rather than the bottom GAN.<br> | ||
While the bottom layer is a purely unconditional GAN, the intermediate generators are more akin to conditional GANs. | While the bottom layer is a purely unconditional GAN, the intermediate generators are more akin to conditional GANs. | ||
===Super-Resolution=== | ===Super-Resolution=== | ||
Upscaling | |||
===Paint-to-Image=== | ===Paint-to-Image=== | ||
Convert a drawing to an image. | |||
===Harmonization=== | ===Harmonization=== | ||
Harmonize, or blend the style of a cut-and-pasted piece of image. | |||
===Editing=== | ===Editing=== | ||
===Single Image Animation=== | ===Single Image Animation=== | ||