8000 GitHub - TheSun00000/spacegan: A GAN based app that generates space images given a drawing of color segments. This project was inspired by NVIDIA gaugan.
[go: up one dir, main page]
More Web Proxy on the site http://driver.im/
Skip to content

A GAN based app that generates space images given a drawing of color segments. This project was inspired by NVIDIA gaugan.

Notifications You must be signed in to change notification settings

TheSun00000/spacegan

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

9 Commits
 
 
 
 
 
 
 
 

Repository files navigation

SpaceGAN:

$ f(x) = \underset{C}{\operatorname{argmax}} P(C_{i} | x) \:\ = \underset{C}{\operatorname{argmax}} {P(Cooks | x), P(Orders | x)} \:\ $

SpaceGAN is a deep learning model that generates space images based on a sketch. It is inspired from the paper: Semantic Image Synthesis with Spatially-Adaptive Normalization [arxiv]


Examples:

Sketch Generated image
Sketch Generated image

Demo:

  1. Start by cloning the this repository:
    git clone https://github.com/TheSun00000/spacegan
  2. Download the generator model generator230.pt from here.
  3. Put the downloaded model in the inference folder.
  4. Run inference/main.py:
    python inference/main.py
  5. You would be able to see a small graphical interface like this

6. Select a color, select a pen size, and start sketching your galaxies:

7. Once finished, click on the button "Done":

About

A GAN based app that generates space images given a drawing of color segments. This project was inspired by NVIDIA gaugan.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published
0