Skip to content
Animating paintings using GANs
Python Jupyter Notebook
Branch: master
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Type Name Latest commit message Commit time
Failed to load latest commit information.
Papers
inputs
models
outputs
reference_materials
training_data
.gitignore
LICENSE
README.md
encoder.py
predictor.py
test.py
test_CAM.py
train.py
util.py
vgg16.py

README.md

painting-animation

Animating paintings using CNN:

1. Motion prediction

Recurrently computed on each frame to predict the next frame.

2. Color transfer

Applied independantly to each frame.

To try on any .png image:

  • put your image in the inputs directory
  • go to your terminal and type:
python test.py --gpu 0 -i ./inputs/your_image.png -o ./outputs  
  • the videos (motion loop, color loop, final video) will be generated in the outputs directory

References

This repository contains mostly code of the following paper:

@article{endo2019animatinglandscape,
  title = {Animating Landscape: Self-Supervised Learning of Decoupled Motion and Appearance for Single-Image Video Synthesis},
  author = {Yuki Endo and Yoshihiro Kanamori and Shigeru Kuriyama},
  journal = {ACM Transactions on Graphics (Proceedings of ACM SIGGRAPH Asia 2019)},
  year = {2019},
  volume = 38,
  number = 6,
  pages = {175:1--175:19}
}
You can’t perform that action at this time.