hytseng0509 / Artediting
Modeling Artistic Workflows for Image Generation and Editing (ECCV 2020)
Stars: ✭ 75
Programming Languages
python
139335 projects - #7 most used programming language
Labels
Projects that are alternatives of or similar to Artediting
Cyclegan Qp
Official PyTorch implementation of "Artist Style Transfer Via Quadratic Potential"
Stars: ✭ 59 (-21.33%)
Mutual labels: gan
Sequentialdata Gan
Tensorflow Implementation of GAN modeling for sequential data
Stars: ✭ 69 (-8%)
Mutual labels: gan
Imagedeblurring
A Keras implementation of image deblurring based on ICCV 2017 paper "Deep Generative Filter for motion deblurring"
Stars: ✭ 72 (-4%)
Mutual labels: gan
Pix2pix
Image-to-image translation with conditional adversarial nets
Stars: ✭ 8,765 (+11586.67%)
Mutual labels: gan
Dcgan Tensorflow
A Tensorflow implementation of Deep Convolutional Generative Adversarial Networks trained on Fashion-MNIST, CIFAR-10, etc.
Stars: ✭ 70 (-6.67%)
Mutual labels: gan
Keras Gan
Keras implementations of Generative Adversarial Networks.
Stars: ✭ 8,494 (+11225.33%)
Mutual labels: gan
Cycle Gan Tf
Reimplementation of cycle-gan(https://arxiv.org/pdf/1703.10593.pdf) with improved w-gan(https://arxiv.org/abs/1704.00028) loss in tensorflow.
Stars: ✭ 74 (-1.33%)
Mutual labels: gan
Pacgan
[NeurIPS 2018] [JSAIT] PacGAN: The power of two samples in generative adversarial networks
Stars: ✭ 67 (-10.67%)
Mutual labels: gan
Frechet Inception Distance
CPU/GPU/TPU implementation of the Fréchet Inception Distance
Stars: ✭ 66 (-12%)
Mutual labels: gan
Geneva
Code to train and evaluate the GeNeVA-GAN model for the GeNeVA task proposed in our ICCV 2019 paper "Tell, Draw, and Repeat: Generating and modifying images based on continual linguistic instruction"
Stars: ✭ 71 (-5.33%)
Mutual labels: gan
Deblurgan Tf
Unofficial tensorflow (tf) implementation of DeblurGAN: Blind Motion Deblurring Using Conditional Adversarial Networks
Stars: ✭ 60 (-20%)
Mutual labels: gan
Hypergan
Composable GAN framework with api and user interface
Stars: ✭ 1,104 (+1372%)
Mutual labels: gan
Colorizer
Add colors to black and white images with neural networks (GANs).
Stars: ✭ 69 (-8%)
Mutual labels: gan
Pytorch Book
PyTorch tutorials and fun projects including neural talk, neural style, poem writing, anime generation (《深度学习框架PyTorch：入门与实战》)
Stars: ✭ 9,546 (+12628%)
Mutual labels: gan
Pytorch Pix2pix
Pytorch implementation of pix2pix for various datasets.
Stars: ✭ 74 (-1.33%)
Mutual labels: gan
Man
Multinomial Adversarial Networks for Multi-Domain Text Classification (NAACL 2018)
Stars: ✭ 72 (-4%)
Mutual labels: gan
Modeling Artistic Workflows for Image Generation and Editing
Pytorch implementation for our artwork generation and editing method. The proposed design can 1) model creation workflows for different types of artwork and 2) enable both the image generation and editing at different workflow stages.
Paper
Please cite our paper if you find the code or dataset useful for your research.
Modeling Artistic Workflows for Image Generation and Editing
Hung-Yu Tseng, Matt Fisher, Jingwan (Cynthia) Lu, Yijun Li, Vladimir (Vova) Kim, Ming-Hsuan Yang
European Conference on Computer Vision (ECCV), 2020
@inproceedings{tseng2020artediting,
author = {Tseng, Hung-Yu and Fisher, Matthew and Lu, Jingwan and Li, Yijun and Kim, Vladimir and Yang, Ming-Hsuan},
booktitle = {European Conference on Computer Vision},
title = {Modeling Artistic Workflows for Image Generation and Editing},
year = {2020}
}
Usage
Installation
Clone this repo:
git clone https://github.com/hytseng0509/ArtEditing
cd ArtEditing
Install packages:
conda create --name artediting python=3.6
conda activate artediting
conda install pytorch torchvision -c pytorch
pip install -r requirements.txt
Datasets
Download the face or anime datasets using the following script:
- Set
DATASETtofaceoranime
cd data
python download_dataset.sh DATASET
cd ..
Workflow Inference Training
We first train the workflow inference model.
python train_inference.py --name face_inference --n_ep_separate 15 --n_ep_joint 15
Artwork Generation Training
Then we load the trained inference model and train the artwork generation model. We need 4 GPUs for batch size of 8.
python train_generation.py --gpu_ids 0,1,2,3 --name face --n_ep_separate 40 --n_ep_joint 15 --load_inference face_inference/30.pth
Learning-Based Regularization Training
Finally, for each workflow stage, we train the regularization for the input image reconstruction.
python train_regularization.py --load face/55.pth --name face_reg0 --reg_stage 0
python train_regularization.py --load face/55.pth --name face_reg1 --reg_stage 1
Testing
Generate reconsturction and random editing results:
python test.py --name face_results --load face/55.pth --reg_load face_reg0/500.pth,face_reg1/500.pth
The results can be found at results/face_results.
Notes
- Part of this implementation is modified from BicycleGAN and MUNIT.
- The dataset, model, and code are for non-commercial research purposes only.
Note that the project description data, including the texts, logos, images, and/or trademarks,
for each open source project belongs to its rightful owner.
If you wish to add or remove any projects, please contact us at [email protected].