StyleCariGAN: Caricature Generation via StyleGAN Feature Map Modulation

This repository contains the official PyTorch implementation of the following paper:

StyleCariGAN: Caricature Generation via StyleGAN Feature Map Modulation
Wonjong Jang, Gwangjin Ju, Yucheol Jung, Jiaolong Yang, Xin Tong, Seungyong Lee, SIGGRAPH 2021

Requirements

PyTorch 1.3.1
torchvision 0.4.2
CUDA 10.1/10.2
dlib 19.22.0
requests 2.23.0
tqdm 4.46.2

If you are using Anaconda environment and get errors regarding compiler version mismatch, check issue #1.

Usage

First download pre-trained model weights:

bash ./download.sh

Train

python -m torch.distributed.launch –nproc_per_node=N_GPU train.py –name EXPERIMENT_NAME –freeze_D

Test

Test on user’s input images:

python test.py –ckpt CHECKPOINT_PATH –input_dir INPUT_IMAGE_PATH –output_dir OUTPUT_CARICATURE_PATH –invert_images

We provide some sample images. Test on sample images:

python test.py –ckpt CHECKPOINT_PATH –input_dir examples/samples –output_dir examples/results –invert_images

It inverts latent codes from input photos and generates caricatures from latent codes.

Input image	Output caricature

Citation

If you find this code useful, please consider citing:

@article{Jang2021StyleCari,
  author    = {Wonjong Jang and Gwangjin Ju and Yucheol Jung and Jiaolong Yang and Xin Tong and Seungyong Lee},
  title     = {StyleCariGAN: Caricature Generation via StyleGAN Feature Map Modulation},
  booktitle = {ACM Transactions on Graphics (Proceedings of ACM SIGGRAPH)},
  publisher = {ACM},
  volume = {40},
  number = {4},
  year = {2021}
}