Encoding in Style: a StyleGAN Encoder for Image-to-Image Translation
We present a generic image-to-image translation framework, pixel2style2pixel (pSp). Our pSp framework is based on a novel encoder network that directly generates a series of style vectors which are fed into a pretrained StyleGAN generator, forming the extended W+ latent space. We first show that our encoder can directly embed real images into W+, with no additional optimization. Next, we propose utilizing our encoder to directly solve image-to-image translation tasks, defining them as encoding problems from some input domain into the latent domain. By deviating from the standard "invert first, edit later" methodology used with previous StyleGAN encoders, our approach can handle a variety of tasks even when the input image is not represented in the StyleGAN domain. We show that solving translation tasks through StyleGAN significantly simplifies the training process, as no adversary is required, has better support for solving tasks without pixel-to-pixel correspondence, and inherently supports multi-modal synthesis via the resampling of styles. Finally, we demonstrate the potential of our framework on a variety of facial image-to-image translation tasks, even when compared to state-of-the-art solutions designed specifically for a single task, and further show that it can be extended beyond the human facial domain.
The proposed pixel2style2pixel framework can be used to solve a wide variety of image-to-image translation tasks. Here we show results of pSp on StyleGAN inversion, multi-modal conditional image synthesis, facial frontalization, inpainting and super-resolution.
Description
Official Implementation of our pSp paper for both training and evaluation. The pSp method extends the StyleGAN model to allow solving different image-to-image translation problems using its encoder.
Recent Updates
2020.10.04
: Initial code release
2020.10.06
: Add pSp toonify model (Thanks to the great work from Doron Adler and Justin Pinkney)!
2021.04.23
: Added several new features:
- Added supported for StyleGANs of different resolutions (e.g., 256, 512, 1024). This can be set using the flag
--output_size
, which is set to 1024 by default. - Added support for the MoCo-Based similarity loss introduced in encoder4editing (Tov et al. 2021). More details are provided below.
2021.07.06
: Added support for training with Weights & Biases. See below for details.
Applications
StyleGAN Encoding
Here, we use pSp to find the latent code of real images in the latent domain of a pretrained StyleGAN generator.
Face Frontalization
In this application we want to generate a front-facing face from a given input image.
Conditional Image Synthesis
Here we wish to generate photo-realistic face images from ambiguous sketch images or segmentation maps. Using style-mixing, we inherently support multi-modal synthesis for a single input.
Super Resolution
Given a low-resolution input image, we generate a corresponding high-resolution image. As this too is an ambiguous task, we can use style-mixing to produce several plausible results.
Getting Started
Prerequisites
- Linux or macOS
- NVIDIA GPU + CUDA CuDNN (CPU may be possible with some modifications, but is not inherently supported)
- Python 2 or 3
Installation
- Clone this repo:
git clone https://github.com/eladrich/pixel2style2pixel.git
cd pixel2style2pixel
- Dependencies:
We recommend running this repository using Anaconda. All dependencies for defining the environment are provided inenvironment/psp_env.yaml
.
Inference Notebook
To help visualize the pSp framework on multiple tasks and to help you get started, we provide a Jupyter notebook found in notebooks/inference_playground.ipynb
that allows one to visualize the various applications of pSp.
The notebook will download the necessary pretrained models and run inference on the images found in notebooks/images
.
For the tasks of conditional image synthesis and super resolution, the notebook also demonstrates pSp's ability to perform multi-modal synthesis using style-mixing.
Pretrained Models
Please download the pre-trained models from the following links. Each pSp model contains the entire pSp architecture, including the encoder and decoder weights.
Path | Description |
---|---|
StyleGAN Inversion | pSp trained with the FFHQ dataset for StyleGAN inversion. |
Face Frontalization | pSp trained with the FFHQ dataset for face frontalization. |
Sketch to Image | pSp trained with the CelebA-HQ dataset for image synthesis from sketches. |
Segmentation to Image | pSp trained with the CelebAMask-HQ dataset for image synthesis from segmentation maps. |
Super Resolution | pSp trained with the CelebA-HQ dataset for super resolution (up to x32 down-sampling). |
Toonify | pSp trained with the FFHQ dataset for toonification using StyleGAN generator from Doron Adler and Justin Pinkney. |
If you wish to use one of the pretrained models for training or inference, you may do so using the flag --checkpoint_path
.
In addition, we provide various auxiliary models needed for training your own pSp model from scratch as well as pretrained models needed for computing the ID metrics reported in the paper.
Path | Description |
---|---|
FFHQ StyleGAN | StyleGAN model pretrained on FFHQ taken from rosinality with 1024x1024 output resolution. |
IR-SE50 Model | Pretrained IR-SE50 model taken from TreB1eN for use in our ID loss during pSp training. |
MoCo ResNet-50 | Pretrained ResNet-50 model trained using MOCOv2 for computing MoCo-based similarity loss on non-facial domains. The model is taken from the official implementation. |
CurricularFace Backbone | Pretrained CurricularFace model taken from HuangYG123 for use in ID similarity metric computation. |
MTCNN | Weights for MTCNN model taken from TreB1eN for use in ID similarity metric computation. (Unpack the tar.gz to extract the 3 model weights.) |
By default, we assume that all auxiliary models are downloaded and saved to the directory pretrained_models
. However, you may use your own paths by changing the necessary values in configs/path_configs.py
.
Training
Preparing your Data
- Currently, we provide support for numerous datasets and experiments (encoding, frontalization, etc.).
- Refer to
configs/paths_config.py
to define the necessary data paths and model paths for training and evaluation. - Refer to
configs/transforms_config.py
for the transforms defined for each dataset/experiment. - Finally, refer to
configs/data_configs.py
for the source/target data paths for the train and test sets as well as the transforms.
- Refer to
- If you wish to experiment with your own dataset, you can simply make the necessary adjustments in
data_configs.py
to define your data paths.transforms_configs.py
to define your own data transforms.
As an example, assume we wish to run encoding using ffhq (dataset_type=ffhq_encode
). We first go to configs/paths_config.py
and define:
dataset_paths = {
'ffhq': '/path/to/ffhq/images256x256'
'celeba_test': '/path/to/CelebAMask-HQ/test_img',
}
The transforms for the experiment are defined in the class EncodeTransforms
in configs/transforms_config.py
.
Finally, in configs/data_configs.py
, we define:
DATASETS = {
'ffhq_encode': {
'transforms': transforms_config.EncodeTransforms,
'train_source_root': dataset_paths['ffhq'],
'train_target_root': dataset_paths['ffhq'],
'test_source_root': dataset_paths['celeba_test'],
'test_target_root': dataset_paths['celeba_test'],
},
}
When defining our datasets, we will take the values in the above dictionary.
Training pSp
The main training script can be found in scripts/train.py
.
Intermediate training results are saved to opts.exp_dir
. This includes checkpoints, train outputs, and test outputs.
Additionally, if you have tensorboard installed, you can visualize tensorboard logs in opts.exp_dir/logs
.
Training the pSp Encoder
python scripts/train.py \
--dataset_type=ffhq_encode \
--exp_dir=/path/to/experiment \
--workers=8 \
--batch_size=8 \
--test_batch_size=8 \
--test_workers=8 \
--val_interval=2500 \
--save_interval=5000 \
--encoder_type=GradualStyleEncoder \
--start_from_latent_avg \
--lpips_lambda=0.8 \
--l2_lambda=1 \
--id_lambda=0.1
Frontalization
python scripts/train.py \
--dataset_type=ffhq_frontalize \
--exp_dir=/path/to/experiment \
--workers=8 \
--batch_size=8 \
--test_batch_size=8 \
--test_workers=8 \
--val_interval=2500 \
--save_interval=5000 \
--encoder_type=GradualStyleEncoder \
--start_from_latent_avg \
--lpips_lambda=0.08 \
--l2_lambda=0.001 \
--lpips_lambda_crop=0.8 \
--l2_lambda_crop=0.01 \
--id_lambda=1 \
--w_norm_lambda=0.005
Sketch to Face
python scripts/train.py \
--dataset_type=celebs_sketch_to_face \
--exp_dir=/path/to/experiment \
--workers=8 \
--batch_size=8 \
--test_batch_size=8 \
--test_workers=8 \
--val_interval=2500 \
--save_interval=5000 \
--encoder_type=GradualStyleEncoder \
--start_from_latent_avg \
--lpips_lambda=0.8 \
--l2_lambda=1 \
--id_lambda=0 \
--w_norm_lambda=0.005 \
--label_nc=1 \
--input_nc=1
Segmentation Map to Face
python scripts/train.py \
--dataset_type=celebs_seg_to_face \
--exp_dir=/path/to/experiment \
--workers=8 \
--batch_size=8 \
--test_batch_size=8 \
--test_workers=8 \
--val_interval=2500 \
--save_interval=5000 \
--encoder_type=GradualStyleEncoder \
--start_from_latent_avg \
--lpips_lambda=0.8 \
--l2_lambda=1 \
--id_lambda=0 \
--w_norm_lambda=0.005 \
--label_nc=19 \
--input_nc=19
Notice with conditional image synthesis no identity loss is utilized (i.e. --id_lambda=0
)
Super Resolution
python scripts/train.py \
--dataset_type=celebs_super_resolution \
--exp_dir=/path/to/experiment \
--workers=8 \
--batch_size=8 \
--test_batch_size=8 \
--test_workers=8 \
--val_interval=2500 \
--save_interval=5000 \
--encoder_type=GradualStyleEncoder \
--start_from_latent_avg \
--lpips_lambda=0.8 \
--l2_lambda=1 \
--id_lambda=0.1 \
--w_norm_lambda=0.005 \
--resize_factors=1,2,4,8,16,32
Additional Notes
- See
options/train_options.py
for all training-specific flags. - See
options/test_options.py
for all test-specific flags. - If you wish to resume from a specific checkpoint (e.g. a pretrained pSp model), you may do so using
--checkpoint_path
. - By default, we assume that the StyleGAN used outputs images at resolution
1024x1024
. If you wish to use a StyleGAN at a smaller resolution, you can do so by using the flag--output_size
(e.g.,--output_size=256
). - If you wish to generate images from segmentation maps, please specify
--label_nc=N
and--input_nc=N
whereN
is the number of semantic categories. - Similarly, for generating images from sketches, please specify
--label_nc=1
and--input_nc=1
. - Specifying
--label_nc=0
(the default value), will directly use the RGB colors as input.
Identity/Similarity Losses
In pSp, we introduce a facial identity loss using a pre-trained ArcFace network for facial recognition. When operating on the human facial domain, we highly recommend employing this loss objective by using the flag --id_lambda
.
In a more recent paper, encoder4editing, the authors generalize this identity loss to other domains by using a MoCo-based ResNet to extract features instead of an ArcFace network. Applying this MoCo-based similarity loss can be done by using the flag --moco_lambda
. We recommend setting --moco_lambda=0.5
in your experiments.
Please note, you cannot set both id_lambda
and moco_lambda
to be active simultaneously (e.g., to use the MoCo-based loss, you should specify, --moco_lambda=0.5 --id_lambda=0
).
Weights & Biases Integration
To help track your experiments, we've integrated Weights & Biases into our training process. To enable Weights & Biases (wandb
), first make an account on the platform's webpage and install wandb
using pip install wandb
. Then, to train pSp using wandb
, simply add the flag --use_wandb
.
Note that when running for the first time, you will be asked to provide your access key which can be accessed via the Weights & Biases platform.
Using Weights & Biases will allow you to visualize the training and testing loss curves as well as intermediate training results.
Testing
Inference
Having trained your model, you can use scripts/inference.py
to apply the model on a set of images.
For example,
python scripts/inference.py \
--exp_dir=/path/to/experiment \
--checkpoint_path=experiment/checkpoints/best_model.pt \
--data_path=/path/to/test_data \
--test_batch_size=4 \
--test_workers=4 \
--couple_outputs
Additional notes to consider:
- During inference, the options used during training are loaded from the saved checkpoint and are then updated using the test options passed to the inference script. For example, there is no need to pass
--dataset_type
or--label_nc
to the inference script, as they are taken from the loadedopts
. - When running inference for segmentation-to-image or sketch-to-image, it is highly recommend to do so with a style-mixing, as is done in the paper. This can simply be done by adding
--latent_mask=8,9,10,11,12,13,14,15,16,17
when calling the script. - When running inference for super-resolution, please provide a single down-sampling value using
--resize_factors
. - Adding the flag
--couple_outputs
will save an additional image containing the input and output images side-by-side in the sub-directoryinference_coupled
. Otherwise, only the output image is saved to the sub-directoryinference_results
. - By default, the images will be saved at resolutiosn of 1024x1024, the original output size of StyleGAN. If you wish to save outputs resized to resolutions of 256x256, you can do so by adding the flag
--resize_outputs
.
Multi-Modal Synthesis with Style-Mixing
Given a trained model for conditional image synthesis or super-resolution, we can easily generate multiple outputs for a given input image. This can be done using the script scripts/style_mixing.py
.
For example, running the following command will perform style-mixing for a segmentation-to-image experiment:
python scripts/style_mixing.py \
--exp_dir=/path/to/experiment \
--checkpoint_path=/path/to/experiment/checkpoints/best_model.pt \
--data_path=/path/to/test_data/ \
--test_batch_size=4 \
--test_workers=4 \
--n_images=25 \
--n_outputs_to_generate=5 \
--latent_mask=8,9,10,11,12,13,14,15,16,17
Here, we inject 5
randomly drawn vectors and perform style-mixing on the latents [8,9,10,11,12,13,14,15,16,17]
.
Additional notes to consider:
- To perform style-mixing on a subset of images, you may use the flag
--n_images
. The default value ofNone
will perform style mixing on every image in the givendata_path
. - You may also include the argument
--mix_alpha=m
wherem
is a float defining the mixing coefficient between the input latent and the randomly drawn latent. - When performing style-mixing for super-resolution, please provide a single down-sampling value using
--resize_factors
. - By default, the images will be saved at resolutiosn of 1024x1024, the original output size of StyleGAN. If you wish to save outputs resized to resolutions of 256x256, you can do so by adding the flag
--resize_outputs
.
Computing Metrics
Similarly, given a trained model and generated outputs, we can compute the loss metrics on a given dataset.
These scripts receive the inference output directory and ground truth directory.
- Calculating the identity loss:
python scripts/calc_id_loss_parallel.py \
--data_path=/path/to/experiment/inference_outputs \
--gt_path=/path/to/test_images \
- Calculating LPIPS loss:
python scripts/calc_losses_on_images.py \
--mode lpips
--data_path=/path/to/experiment/inference_outputs \
--gt_path=/path/to/test_images \
- Calculating L2 loss:
python scripts/calc_losses_on_images.py \
--mode l2
--data_path=/path/to/experiment/inference_outputs \
--gt_path=/path/to/test_images \
Additional Applications
To better show the flexibility of our pSp framework we present additional applications below.
As with our main applications, you may download the pretrained models here:
Path | Description |
---|---|
Toonify | pSp trained with the FFHQ dataset for toonification using StyleGAN generator from Doron Adler and Justin Pinkney. |
Toonify
Using the toonify StyleGAN built by Doron Adler and Justin Pinkney, we take a real face image and generate a toonified version of the given image. We train the pSp encoder to directly reconstruct real face images inside the toons latent space resulting in a projection of each image to the closest toon. We do so without requiring any labeled pairs or distillation!
This is trained exactly like the StyleGAN inversion task with several changes:
- Change from FFHQ StyleGAN to toonifed StyleGAN (can be set using
--stylegan_weights
)- The toonify generator is taken from Doron Adler and Justin Pinkney and converted to Pytorch using rosinality's conversion script.
- For convenience, the converted generator Pytorch model may be downloaded here.
- Increase
id_lambda
from0.1
to1
- Increase
w_norm_lambda
from0.005
to0.025
We obtain the best results after around 6000
iterations of training (can be set using --max_steps
)
Repository structure
Path | Description |
---|---|
pixel2style2pixel | Repository root folder |
├ configs | Folder containing configs defining model/data paths and data transforms |
├ criteria | Folder containing various loss criterias for training |
├ datasets | Folder with various dataset objects and augmentations |
├ environment | Folder containing Anaconda environment used in our experiments |
├ models | Folder containting all the models and training objects |
│ ├ encoders | Folder containing our pSp encoder architecture implementation and ArcFace encoder implementation from TreB1eN |
│ ├ mtcnn | MTCNN implementation from TreB1eN |
│ ├ stylegan2 | StyleGAN2 model from rosinality |
│ └ psp.py | Implementation of our pSp framework |
├ notebook | Folder with jupyter notebook containing pSp inference playground |
├ options | Folder with training and test command-line options |
├ scripts | Folder with running scripts for training and inference |
├ training | Folder with main training logic and Ranger implementation from lessw2020 |
├ utils | Folder with various utility functions |
TODOs
- Add multi-gpu support
Credits
StyleGAN2 implementation:
https://github.com/rosinality/stylegan2-pytorch
Copyright (c) 2019 Kim Seonghyeon
License (MIT) https://github.com/rosinality/stylegan2-pytorch/blob/master/LICENSE
MTCNN, IR-SE50, and ArcFace models and implementations:
https://github.com/TreB1eN/InsightFace_Pytorch
Copyright (c) 2018 TreB1eN
License (MIT) https://github.com/TreB1eN/InsightFace_Pytorch/blob/master/LICENSE
CurricularFace model and implementation:
https://github.com/HuangYG123/CurricularFace
Copyright (c) 2020 HuangYG123
License (MIT) https://github.com/HuangYG123/CurricularFace/blob/master/LICENSE
Ranger optimizer implementation:
https://github.com/lessw2020/Ranger-Deep-Learning-Optimizer
License (Apache License 2.0) https://github.com/lessw2020/Ranger-Deep-Learning-Optimizer/blob/master/LICENSE
LPIPS implementation:
https://github.com/S-aiueo32/lpips-pytorch
Copyright (c) 2020, Sou Uchida
License (BSD 2-Clause) https://github.com/S-aiueo32/lpips-pytorch/blob/master/LICENSE
Please Note: The CUDA files under the StyleGAN2 ops directory are made available under the Nvidia Source Code License-NC
Inspired by pSp
Below are several works inspired by pSp that we found particularly interesting:
Reverse Toonification
Using our pSp encoder, artist Nathan Shipley transformed animated figures and paintings into real life. Check out his amazing work on his twitter page and website.
Deploying pSp with StyleSpace for Editing
Awesome work from Justin Pinkney who deployed our pSp model on Runway and provided support for editing the resulting inversions using the StyleSpace Analysis paper. Check out his repository here.
Encoder4Editing (e4e)
Building on the work of pSp, Tov et al. design an encoder to enable high quality edits on real images. Check out their paper and code.
Style-based Age Manipulation (SAM)
Leveraging pSp and the rich semantics of StyleGAN, SAM learns non-linear latent space paths for modeling the age transformation of real face images. Check out the project page here.
ReStyle
ReStyle builds on recent encoders such as pSp and e4e by introducing an iterative refinment mechanism to gradually improve the inversion of real images. Check out the project page here.
pSp in the Media
- bycloud: AI Generates Cartoon Characters In Real Life Pixel2Style2Pixel
- Synced: Pixel2Style2Pixel: Novel Encoder Architecture Boosts Facial Image-To-Image Translation
- Cartoon Brew: An Artist Has Used Machine Learning To Turn Animated Characters Into Creepy Photorealistic Figures
Citation
If you use this code for your research, please cite our paper Encoding in Style: a StyleGAN Encoder for Image-to-Image Translation:
@InProceedings{richardson2021encoding,
author = {Richardson, Elad and Alaluf, Yuval and Patashnik, Or and Nitzan, Yotam and Azar, Yaniv and Shapiro, Stav and Cohen-Or, Daniel},
title = {Encoding in Style: a StyleGAN Encoder for Image-to-Image Translation},
booktitle = {IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)},
month = {June},
year = {2021}
}