SG Adaption

• macos tiff to png: sips -s format png *.tiff --out .

Stylegan2 Align, Project, Animate, Mix Styles and Train

This is a fork of stylegan2, with some easy-to-use util functions to generate animations and mix styles.

This is a combination of code from following repos:

• NVlabs/stylegan2: original repo
• rolux/stylegan2encoder: modified projector
• ak9250/stylegan2: training stylegan2
• iyaja/stylegan-encoder: stylemix
• ShenYujun/InterFaceGAN: linear interpolation logic

Latent direction trained data taken from here (stylegan2directions):

• https://hostb.org/NCM provided by robertluxemburg

Here is a detailed overview on stylegan-v1, but many suggestions apply to stylegan2 too:

• https://www.gwern.net/Faces#data-preparation

Downloading models:

Sometimes, you might run out of quota while downloading models from gdrive. In such cases, you can manually download the following models and place them in models directory of your project:

stylegan2
|--models
|  |--shape_predictor_68_face_landmarks.dat
|  |--stylegan2-ffhq-config-f.pkl
|  |--vgg16_zhang_perceptual.pkl
|  |--inception_v3_features.pkl # for training only

Download Links:

• Face Landmarks Predictor
• Stylegan2 Model - config f
• VGG16 Perceptual
• Inception V3 model

Generating latent representation of your images, using the modified projector.

You can generate latent representations of your own images using two scripts:

0. Copy images with faces to raw_images/

1. Extract and align faces from images

   python align_images.py raw_images/ aligned_images/

2. Generating latent representation of your images

   python project_images.py aligned_images/ face_datasets/generated_images/

3. Generate transitions from generated_images

   python transition.py face_datasets/generated_images/

   

4. Generate animations from generated_images

• Download stylegan2directions from here and extract to the root folder of your project.

• [Optional] Download face_datasets from here and extract to the root folder of your project. (Image Copyright belongs to the original owners, I got them from Google/Pinterest.)

• Instead of 'face_datasets', you can generate some faces using random latents.

  python run_generator.py generate-images --network=gdrive:networks/stylegan2-ffhq-config-f.pkl --seeds=6600-6625 --truncation-psi=0.5

  Some points:

  • Smile works best, blink is ok, yes and no are bad.
  • I find that the face moves or the structure changes for yes and no and to a lesser extent for blink.
  • How to generate these animations without these artifacts?

  python animate.py --in-file=face_datasets/jdepp/4_01.npy --mode=smile

  

  python animate.py --in-file=face_datasets/jdepp/4_01.npy --mode=blink

  

  python animate.py --in-file=face_datasets/jdepp/4_01.npy --mode=yes

  

  python animate.py --in-file=face_datasets/jdepp/4_01.npy --mode=no

  

  Or you can customise your options. exec '>python animate.py -h' for more examples.

  python animate.py --in-file=face_datasets/jdepp/4_01.npy --dir-file=stylegan2directions/smile.npy --start=-5 --stop=15 --repeat 2 --steps=10 --reverse

  Some samples of age and gender interpolation:

  Age:

  

  Gender:

  

5. Do some style mixing, you can either mix a persons expression or mix the makeup.

   The style mixing here is different from the original repo. I have used the technique from - iyaja/stylegan-encoder

   Instead of using the average latent and truncation, here, the latents from the rows and columns are mixed in 'col-ratio'. Here the latents are [18, 512]. Only the column latents from col-styles are mixed.

   Makeup mix works better. Expression distorts face - resultant face neither looks like a column face nor a row face. How to fix this?

   python run_style_mixer.py --mode=makeup --row-imgs-dir=face_datasets/jdepp/ --col-imgs-dir=face_datasets/avatars/

   python run_style_mixer.py --mode=makeup --row-imgs-dir=face_datasets/avatars/ --col-imgs-dir=face_datasets/avatars/

   python run_style_mixer.py --mode=expression --row-imgs-dir=face_datasets/jdepp/ --col-imgs-dir=face_datasets/mrbean/

   Or you can customise your options. Try 'python run_style_mixer.py -h' for options.

   python run_style_mixer.py --row-imgs-dir=avatars --col-imgs-dir=avatars --col-styles='0-6' --col-ratio=0.7

6. After mixing styles, the resultant latents are stored in the results folder. You can use these to further generate the transition video. You will have to handpick some npy files from results and create a new folder and copy them there and then run the transition.py script.

Original Readme

StyleGAN2 — Official TensorFlow Implementation

Analyzing and Improving the Image Quality of StyleGAN
Tero Karras, Samuli Laine, Miika Aittala, Janne Hellsten, Jaakko Lehtinen, Timo Aila

Paper: http://arxiv.org/abs/1912.04958
Video: https://youtu.be/c-NJtV9Jvp0

Abstract: The style-based GAN architecture (StyleGAN) yields state-of-the-art results in data-driven unconditional generative image modeling. We expose and analyze several of its characteristic artifacts, and propose changes in both model architecture and training methods to address them. In particular, we redesign generator normalization, revisit progressive growing, and regularize the generator to encourage good conditioning in the mapping from latent vectors to images. In addition to improving image quality, this path length regularizer yields the additional benefit that the generator becomes significantly easier to invert. This makes it possible to reliably detect if an image is generated by a particular network. We furthermore visualize how well the generator utilizes its output resolution, and identify a capacity problem, motivating us to train larger models for additional quality improvements. Overall, our improved model redefines the state of the art in unconditional image modeling, both in terms of existing distribution quality metrics as well as perceived image quality.

For business inquiries, please contact researchinquiries@nvidia.com
For press and other inquiries, please contact Hector Marinez at hmarinez@nvidia.com

Additional material
StyleGAN2	Main Google Drive folder
├  stylegan2-paper.pdf	High-quality version of the paper
├  stylegan2-video.mp4	High-quality version of the video
├  images	Example images produced using our method
│  ├  curated-images	Hand-picked images showcasing our results
│  └  100k-generated-images	Random images with and without truncation
├  videos	Individual clips of the video as high-quality MP4
└  networks	Pre-trained networks
├  stylegan2-ffhq-config-f.pkl	StyleGAN2 for FFHQ dataset at 1024×1024
├  stylegan2-car-config-f.pkl	StyleGAN2 for LSUN Car dataset at 512×384
├  stylegan2-cat-config-f.pkl	StyleGAN2 for LSUN Cat dataset at 256×256
├  stylegan2-church-config-f.pkl	StyleGAN2 for LSUN Church dataset at 256×256
├  stylegan2-horse-config-f.pkl	StyleGAN2 for LSUN Horse dataset at 256×256
└ ⋯	Other training configurations used in the paper

Requirements

• Both Linux and Windows are supported. Linux is recommended for performance and compatibility reasons.
• 64-bit Python 3.6 installation. We recommend Anaconda3 with numpy 1.14.3 or newer.
• TensorFlow 1.15 with GPU support. The code does not support TensorFlow 2.0.
• One or more high-end NVIDIA GPUs, NVIDIA drivers, CUDA 10.0 toolkit and cuDNN 7.5. To reproduce the results reported in the paper, you need an NVIDIA GPU with at least 16 GB of DRAM.
• Docker users: use the provided Dockerfile to build an image with the required library dependencies.

StyleGAN2 relies on custom TensorFlow ops that are compiled on the fly using NVCC. To test that your NVCC installation is working correctly, run:

nvcc test_nvcc.cu -o test_nvcc -run
| CPU says hello.
| GPU says hello.

On Windows, the compilation requires Microsoft Visual Studio to be in PATH. We recommend installing Visual Studio Community Edition and adding into PATH using "C:\Program Files (x86)\Microsoft Visual Studio\2019\Community\VC\Auxiliary\Build\vcvars64.bat".

Preparing datasets

Datasets are stored as multi-resolution TFRecords, similar to the original StyleGAN. Each dataset consists of multiple *.tfrecords files stored under a common directory, e.g., ~/datasets/ffhq/ffhq-r*.tfrecords. In the following sections, the datasets are referenced using a combination of --dataset and --data-dir arguments, e.g., --dataset=ffhq --data-dir=~/datasets.

FFHQ. To download the Flickr-Faces-HQ dataset as multi-resolution TFRecords, run:

pushd ~
git clone https://github.com/NVlabs/ffhq-dataset.git
cd ffhq-dataset
python download_ffhq.py --tfrecords
popd
python dataset_tool.py display ~/ffhq-dataset/tfrecords/ffhq

LSUN. Download the desired LSUN categories in LMDB format from the LSUN project page. To convert the data to multi-resolution TFRecords, run:

python dataset_tool.py create_lsun_wide ~/datasets/car ~/lsun/car_lmdb --width=512 --height=384
python dataset_tool.py create_lsun ~/datasets/cat ~/lsun/cat_lmdb --resolution=256
python dataset_tool.py create_lsun ~/datasets/church ~/lsun/church_outdoor_train_lmdb --resolution=256
python dataset_tool.py create_lsun ~/datasets/horse ~/lsun/horse_lmdb --resolution=256

Custom. Create custom datasets by placing all training images under a single directory. The images must be square-shaped and they must all have the same power-of-two dimensions. To convert the images to multi-resolution TFRecords, run:

python dataset_tool.py create_from_images ~/datasets/my-custom-dataset ~/my-custom-images
python dataset_tool.py display ~/datasets/my-custom-dataset

Using pre-trained networks

Pre-trained networks are stored as *.pkl files on the StyleGAN2 Google Drive folder. Below, you can either reference them directly using the syntax gdrive:networks/<filename>.pkl, or download them manually and reference by filename.

Generating images:

# Generate uncurated ffhq images (matches paper Figure 12)
python run_generator.py generate-images --network=gdrive:networks/stylegan2-ffhq-config-f.pkl \
  --seeds=6600-6625 --truncation-psi=0.5

# Generate curated ffhq images (matches paper Figure 11)
python run_generator.py generate-images --network=gdrive:networks/stylegan2-ffhq-config-f.pkl \
  --seeds=66,230,389,1518 --truncation-psi=1.0

# Generate uncurated car images
python run_generator.py generate-images --network=gdrive:networks/stylegan2-car-config-f.pkl \
  --seeds=6000-6025 --truncation-psi=0.5

# Example of style mixing (matches the corresponding video clip)
python run_generator.py style-mixing-example --network=gdrive:networks/stylegan2-ffhq-config-f.pkl \
  --row-seeds=85,100,75,458,1500 --col-seeds=55,821,1789,293 --truncation-psi=1.0

The results are placed in results/<RUNNING_ID>/*.png. You can change the location with --result-dir. For example, --result-dir=~/my-stylegan2-results.

Projecting images to latent space:

# Project generated images
python run_projector.py project-generated-images --network=gdrive:networks/stylegan2-car-config-f.pkl \
  --seeds=0,1,5

# Project real images
python run_projector.py project-real-images --network=gdrive:networks/stylegan2-car-config-f.pkl \
  --dataset=car --data-dir=~/datasets

You can import the networks in your own Python code using pickle.load(). For this to work, you need to include the dnnlib source directory in PYTHONPATH and create a default TensorFlow session by calling dnnlib.tflib.init_tf(). See run_generator.py and pretrained_networks.py for examples.

Training networks

To reproduce the training runs for config F in Tables 1 and 3, run:

python run_training.py --num-gpus=8 --data-dir=~/datasets --config=config-f \
  --dataset=ffhq --mirror-augment=true
python run_training.py --num-gpus=8 --data-dir=~/datasets --config=config-f \
  --dataset=car --total-kimg=57000
python run_training.py --num-gpus=8 --data-dir=~/datasets --config=config-f \
  --dataset=cat --total-kimg=88000
python run_training.py --num-gpus=8 --data-dir=~/datasets --config=config-f \
  --dataset=church --total-kimg 88000 --gamma=100
python run_training.py --num-gpus=8 --data-dir=~/datasets --config=config-f \
  --dataset=horse --total-kimg 100000 --gamma=100

For other configurations, see python run_training.py --help.

We have verified that the results match the paper when training with 1, 2, 4, or 8 GPUs. Note that training FFHQ at 1024×1024 resolution requires GPU(s) with at least 16 GB of memory. The following table lists typical training times using NVIDIA DGX-1 with 8 Tesla V100 GPUs:

Configuration	Resolution	Total kimg	1 GPU	2 GPUs	4 GPUs	8 GPUs	GPU mem
config-f	1024×1024	25000	69d 23h	36d 4h	18d 14h	9d 18h	13.3 GB
config-f	1024×1024	10000	27d 23h	14d 11h	7d 10h	3d 22h	13.3 GB
config-e	1024×1024	25000	35d 11h	18d 15h	9d 15h	5d 6h	8.6 GB
config-e	1024×1024	10000	14d 4h	7d 11h	3d 20h	2d 3h	8.6 GB
config-f	256×256	25000	32d 13h	16d 23h	8d 21h	4d 18h	6.4 GB
config-f	256×256	10000	13d 0h	6d 19h	3d 13h	1d 22h	6.4 GB

Training curves for FFHQ config F (StyleGAN2) compared to original StyleGAN using 8 GPUs:

After training, the resulting networks can be used the same way as the official pre-trained networks:

# Generate 1000 random images without truncation
python run_generator.py generate-images --seeds=0-999 --truncation-psi=1.0 \
  --network=results/00006-stylegan2-ffhq-8gpu-config-f/networks-final.pkl

Evaluation metrics

To reproduce the numbers for config F in Tables 1 and 3, run:

python run_metrics.py --data-dir=~/datasets --network=gdrive:networks/stylegan2-ffhq-config-f.pkl \
  --metrics=fid50k,ppl_wend --dataset=ffhq --mirror-augment=true
python run_metrics.py --data-dir=~/datasets --network=gdrive:networks/stylegan2-car-config-f.pkl \
  --metrics=fid50k,ppl2_wend --dataset=car
python run_metrics.py --data-dir=~/datasets --network=gdrive:networks/stylegan2-cat-config-f.pkl \
  --metrics=fid50k,ppl2_wend --dataset=cat
python run_metrics.py --data-dir=~/datasets --network=gdrive:networks/stylegan2-church-config-f.pkl \
  --metrics=fid50k,ppl2_wend --dataset=church
python run_metrics.py --data-dir=~/datasets --network=gdrive:networks/stylegan2-horse-config-f.pkl \
  --metrics=fid50k,ppl2_wend --dataset=horse

For other configurations, see the StyleGAN2 Google Drive folder.

Note that the metrics are evaluated using a different random seed each time, so the results will vary between runs. In the paper, we reported the average result of running each metric 10 times. The following table lists the available metrics along with their expected runtimes and random variation:

Metric	FFHQ config F	1 GPU	2 GPUs	4 GPUs	Description
fid50k	2.84 ± 0.03	22 min	14 min	10 min	Fréchet Inception Distance
is50k	5.13 ± 0.02	23 min	14 min	8 min	Inception Score
ppl_zfull	348.0 ± 3.8	41 min	22 min	14 min	Perceptual Path Length in Z, full paths
ppl_wfull	126.9 ± 0.2	42 min	22 min	13 min	Perceptual Path Length in W, full paths
ppl_zend	348.6 ± 3.0	41 min	22 min	14 min	Perceptual Path Length in Z, path endpoints
ppl_wend	129.4 ± 0.8	40 min	23 min	13 min	Perceptual Path Length in W, path endpoints
ppl2_wend	145.0 ± 0.5	41 min	23 min	14 min	Perceptual Path Length without center crop
ls	154.2 / 4.27	10 hrs	6 hrs	4 hrs	Linear Separability
pr50k3	0.689 / 0.492	26 min	17 min	12 min	Precision and Recall

Note that some of the metrics cache dataset-specific data on the disk, and they will take somewhat longer when run for the first time.

License

Copyright © 2019, NVIDIA Corporation. All rights reserved.

This work is made available under the Nvidia Source Code License-NC. To view a copy of this license, visit https://nvlabs.github.io/stylegan2/license.html

Citation

@article{Karras2019stylegan2,
  title   = {Analyzing and Improving the Image Quality of {StyleGAN}},
  author  = {Tero Karras and Samuli Laine and Miika Aittala and Janne Hellsten and Jaakko Lehtinen and Timo Aila},
  journal = {CoRR},
  volume  = {abs/1912.04958},
  year    = {2019},
}

Acknowledgements

We thank Ming-Yu Liu for an early review, Timo Viitanen for his help with code release, and Tero Kuosmanen for compute infrastructure.