mimicry

by kwotsin

kwotsin / mimicry

[CVPR 2020 Workshop] A PyTorch GAN library that reproduces research results for popular GANs.

405 Stars 38 Forks Last release: 2 months ago (0.1.16) MIT License 152 Commits 3 Releases

Available items

No Items, yet!

The developer of this repository has not created any items for sale yet. Need a bug fixed? Help with integration? A different license? Create a request here:

alt text


CircleCI codecov PyPI version Documentation Status License: MIT

About | Documentation | Tutorial | Gallery | Paper

Mimicry is a lightweight PyTorch library aimed towards the reproducibility of GAN research.

Comparing GANs is often difficult - mild differences in implementations and evaluation methodologies can result in huge performance differences. Mimicry aims to resolve this by providing: (a) Standardized implementations of popular GANs that closely reproduce reported scores; (b) Baseline scores of GANs trained and evaluated under the same conditions; (c) A framework for researchers to focus on implementation of GANs without rewriting most of GAN training boilerplate code, with support for multiple GAN evaluation metrics.

We provide a model zoo and set of baselines to benchmark different GANs of the same model size trained under the same conditions, using multiple metrics. To ensure reproducibility, we verify scores of our implemented models against reported scores in literature.


Installation

The library can be installed with:

pip install torch-mimicry
To install the latest master version instead:
pip install git+https://github.com/kwotsin/mimicry.git

See also setup information for more.

Example Usage

Training a popular GAN like SNGAN that reproduces reported scores can be done as simply as: ```python import torch import torch.optim as optim import torchmimicry as mmc from torchmimicry.nets import sngan

Data handling objects

device = torch.device('cuda:0' if torch.cuda.isavailable() else "cpu") dataset = mmc.datasets.loaddataset(root='./datasets', name='cifar10') dataloader = torch.utils.data.DataLoader( dataset, batchsize=64, shuffle=True, numworkers=4)

Define models and optimizers

netG = sngan.SNGANGenerator32().to(device) netD = sngan.SNGANDiscriminator32().to(device) optD = optim.Adam(netD.parameters(), 2e-4, betas=(0.0, 0.9)) optG = optim.Adam(netG.parameters(), 2e-4, betas=(0.0, 0.9))

Start training

trainer = mmc.training.Trainer( netD=netD, netG=netG, optD=optD, optG=optG, ndis=5, numsteps=100000, lrdecay='linear', dataloader=dataloader, logdir='./log/example', device=device) trainer.train()

Evaluate fid

mmc.metrics.evaluate( metric='fid', logdir='./log/example', netG=netG, datasetname='cifar10', numrealsamples=50000, numfakesamples=50000, evaluate_step=100000, device=device)

Example outputs:

INFO: [Epoch 1/127][Global Step: 10/100000] | D(G(z)): 0.5941 | D(x): 0.9303 | errD: 1.4052 | errG: -0.6671 | lrD: 0.0002 | lrG: 0.0002 | (0.4550 sec/idx) ^CINFO: Saving checkpoints from keyboard interrupt... INFO: Training Ended

Tensorboard visualizations:
tensorboard --logdir=./log/example ``` See further details in example script, as well as a detailed tutorial on implementing a custom GAN from scratch.

Further Guides

Baselines | Model Zoo

For a fair comparison, we train all models under the same training conditions for each dataset, each implemented using ResNet backbones of the same architectural capacity. We train our models with the Adam optimizer using the popular hyperparameters (β1, β2) = (0.0, 0.9). ndis represents the number of discriminator update steps per generator update step, and niter is simply the number of training iterations.

Models

| Abbrev. | Name | Type* | |:-----------:|:---------------------------------------------:|:-------------:| | DCGAN | Deep Convolutional GAN | Unconditional | | WGAN-GP | Wasserstein GAN with Gradient Penalty | Unconditional | | SNGAN | Spectral Normalization GAN | Unconditional | | cGAN-PD | Conditional GAN with Projection Discriminator | Conditional | | SSGAN | Self-supervised GAN | Unconditional | | InfoMax-GAN | Infomax-GAN | Unconditional |

*Conditional GAN scores are only reported for labelled datasets.

Metrics

| Metric | Method | |:--------------------------------:|:---------------------------------------:| | Inception Score (IS)* | 50K samples at 10 splits| | Fréchet Inception Distance (FID) | 50K real/generated samples | | Kernel Inception Distance (KID) | 50K real/generated samples, averaged over 10 splits.|

*Inception Score can be a poor indicator of GAN performance, as it does not measure diversity and is not domain agnostic. This is why certain datasets with only a single class (e.g. CelebA and LSUN-Bedroom) will perform poorly when using this metric.

Datasets

| Dataset | Split | Resolution | |:------------:|:---------:|:----------:| | CIFAR-10 | Train | 32 x 32 | | CIFAR-100 | Train | 32 x 32 | | ImageNet | Train | 32 x 32 | | STL-10 | Unlabeled | 48 x 48 | | CelebA | All | 64 x 64 | | CelebA | All | 128 x 128 | | LSUN-Bedroom | Train | 128 x 128 | | ImageNet | Train | 128 x 128 |


CelebA

Paper | Dataset

Training Parameters

| Resolution | Batch Size | Learning Rate | β1 | β2 | Decay Policy | ndis | niter | |:----------:|:----------:|:-------------:|:-------------:|:-------------:|:------------:|:---------------:|------------------| | 128 x 128 | 64 | 2e-4 | 0.0 | 0.9 | None | 2 | 100K | | 64 x 64 | 64 | 2e-4 | 0.0 | 0.9 | Linear | 5 | 100K |

Results

| Resolution | Model | IS | FID | KID | Checkpoint | Code | |:---:|:---:|:---:|:---:|:---:|:---:|:---:| | 128 x 128 | SNGAN | 2.72 ± 0.01 | 12.93 ± 0.04 | 0.0076 ± 0.0001 | netG.pth | sngan_128.py | | 128 x 128 | SSGAN | 2.63 ± 0.01 | 15.18 ± 0.10 | 0.0101 ± 0.0001 | netG.pth | ssgan_128.py | | 128 x 128 | InfoMax-GAN | 2.84 ± 0.01 | 9.50 ± 0.04 | 0.0063 ± 0.0001 | netG.pth | infomaxgan128.py | | 64 x 64 | SNGAN | 2.68 ± 0.01 | 5.71 ± 0.02 | 0.0033 ± 0.0001 | netG.pth | sngan_64.py | | 64 x 64 | SSGAN | 2.67 ± 0.01 | 6.03 ± 0.04 | 0.0036 ± 0.0001 | netG.pth | ssgan_64.py | | 64 x 64 | InfoMax-GAN |2.68 ± 0.01 | 5.71 ± 0.06 | 0.0033 ± 0.0001 | netG.pth | infomaxgan64.py |

LSUN-Bedroom

Paper | Dataset

Training Parameters

| Resolution | Batch Size | Learning Rate | β1 | β2 | Decay Policy | ndis | niter | |:----------:|:----------:|:-------------:|:-------------:|:-------------:|:------------:|:---------------:|------------------| | 128 x 128 | 64 | 2e-4 | 0.0 | 0.9 | Linear | 2 | 100K |

Results

| Resolution | Model | IS | FID | KID | Checkpoint | Code | |:---:|:---:|:---:|:---:|:---:|:---:|:---:| | 128 x 128 | SNGAN | 2.30 ± 0.01 | 25.87 ± 0.03 | 0.0141 ± 0.0001 | netG.pth | sngan_128.py | | 128 x 128 | SSGAN | 2.12 ± 0.01 | 12.02 ± 0.07 | 0.0077 ± 0.0001 | netG.pth | ssgan_128.py | | 128 x 128 | InfoMax-GAN |2.22 ± 0.01 | 12.13 ± 0.16 | 0.0080 ± 0.0001 | netG.pth | infomaxgan128.py |

STL-10

Paper | Dataset

Training Parameters

| Resolution | Batch Size | Learning Rate | β1 | β2 | Decay Policy | ndis | niter | |:----------:|:----------:|:-------------:|:-------------:|:-------------:|:------------:|:---------------:|------------------| | 48 x 48 | 64 | 2e-4 | 0.0 | 0.9 | Linear | 5 | 100K |

Results

| Resolution | Model | IS | FID | KID | Checkpoint | Code | |:---:|:---:|:---:|:---:|:---:|:---:|:---:| | 48 x 48 | WGAN-GP | 8.55 ± 0.02 | 43.01 ± 0.19 | 0.0440 ± 0.0003 | netG.pth | wgangp48.py | | 48 x 48 | SNGAN | 8.04 ± 0.07 | 39.56 ± 0.10 | 0.0369 ± 0.0002 | netG.pth | sngan_48.py | | 48 x 48 | SSGAN | 8.25 ± 0.06 | 37.06 ± 0.19 | 0.0332 ± 0.0004| netG.pth | ssgan_48.py | | 48 x 48 | InfoMax-GAN | 8.54 ± 0.12 | 35.52 ± 0.10 | 0.0326 ± 0.0002 | netG.pth | infomaxgan48.py |

ImageNet

Paper | Dataset

Training Parameters

| Resolution | Batch Size | Learning Rate | β1 | β2 | Decay Policy | ndis | niter | |:----------:|:----------:|:-------------:|:-------------:|:-------------:|:------------:|:---------------:|------------------| | 32 x 32 | 64 | 2e-4 | 0.0 | 0.9 | Linear | 5 | 100K | | 128 x 128 | 64 | 2e-4 | 0.0 | 0.9 | None | 5 | 450k |

Results

| Resolution | Model | IS | FID | KID | Checkpoint | Code | |:---:|:---:|:---:|:---:|:---:|:---:|:---:| | 128 x 128 | SNGAN | 13.05 ± 0.05 | 65.74 ± 0.31 | 0.0663 ± 0.0004 | netG.pth | sngan_128.py | | 128 x 128 | SSGAN | 13.30 ± 0.03 | 62.48 ± 0.31 | 0.0616 ± 0.0004 | netG.pth | ssgan_128.py | | 128 x 128 | InfoMax-GAN | 13.68 ± 0.06 | 58.91 ± 0.14 | 0.0579 ± 0.0004 | netG.pth | infomaxgan128.py | | 32 x 32 | SNGAN | 8.97 ± 0.12 | 23.04 ± 0.06 | 0.0157 ± 0.0002 | netG.pth | sngan_32.py | | 32 x 32 | cGAN-PD | 9.08 ± 0.17 | 21.17 ± 0.05 | 0.0145 ± 0.0002 | netG.pth | cganpd32.py | | 32 x 32 | SSGAN | 9.11 ± 0.12 | 21.79 ± 0.09 | 0.0152 ± 0.0002 | netG.pth | ssgan_32.py | | 32 x 32 | InfoMax-GAN | 9.04 ± 0.10 | 20.68 ± 0.02 | 0.0149 ± 0.0001 | netG.pth | infomaxgan32.py |

CIFAR-10

Paper | Dataset

Training Parameters

| Resolution | Batch Size | Learning Rate | β1 | β2 | Decay Policy | ndis | niter | |:----------:|:----------:|:-------------:|:-------------:|:-------------:|:------------:|:---------------:|------------------| | 32 x 32 | 64 | 2e-4 | 0.0 | 0.9 | Linear | 5 | 100K |

Results

| Resolution | Model | IS | FID | KID | Checkpoint | Code | |:---:|:---:|:---:|:---:|:---:|:---:|:---:| | 32 x 32 | WGAN-GP | 7.33 ± 0.02 | 22.29 ± 0.06 | 0.0204± 0.0004 | netG.pth | wgangp32.py | | 32 x 32 | SNGAN | 7.97 ± 0.06 | 16.77 ± 0.04 | 0.0125 ± 0.0001 | netG.pth | sngan_32.py | | 32 x 32 | cGAN-PD | 8.25 ± 0.13 | 10.84 ± 0.03 | 0.0070 ± 0.0001 | netG.pth | cganpd32.py | | 32 x 32 | SSGAN | 8.17 ± 0.06 | 14.65 ± 0.04 | 0.0101 ± 0.0002 | netG.pth | ssgan_32.py | | 32 x 32 | InfoMax-GAN | 8.08± 0.08 | 15.12 ± 0.10 | 0.0112 ± 0.0001 | netG.pth | infomaxgan32.py |

CIFAR-100

Paper | Dataset

Training Parameters

| Resolution | Batch Size | Learning Rate | β1 | β2 | Decay Policy | ndis | niter | |:----------:|:----------:|:-------------:|:-------------:|:-------------:|:------------:|:---------------:|------------------| | 32 x 32 | 64 | 2e-4 | 0.0 | 0.9 | Linear | 5 | 100K |

Results

| Resolution | Model | IS | FID | KID | Checkpoint | Code | |:---:|:---:|:---:|:---:|:---:|:---:|:---:| | 32 x 32 | SNGAN | 7.57 ± 0.11 | 22.61 ± 0.06 | 0.0156 ± 0.0003 | netG.pth | sngan_32.py | | 32 x 32 | cGAN-PD | 8.92 ± 0.07 | 14.16 ± 0.01 | 0.0085 ± 0.0002 | netG.pth | cganpd32.py | | 32 x 32 | SSGAN | 7.56 ± 0.07 | 22.18 ± 0.10 | 0.0161 ± 0.0002 | netG.pth | ssgan_32.py | | 32 x 32 | InfoMax-GAN | 7.86 ± 0.10 | 18.94 ± 0.13 | 0.0135 ± 0.0004 | netG.pth | infomaxgan32.py |


Reproducibility

To verify our implementations, we reproduce reported scores in literature by re-implementing the models with the same architecture, training them under the same conditions and evaluate them on CIFAR-10 using the exact same methodology for computing FID.

As FID produces highly biased estimates (where using larger samples lead to a lower score), we reproduce the scores using the same sample sizes, where nreal and nfake refers to the number of real and fake images used respectively for computing FID.

| Metric | Model | Score | Reported Score | nreal| nfake| Checkpoint | Code | |:-------------------:|:-----------:|:---------------:|:--------------:|:------------------:|:----:|:----:|:---:| | FID | DCGAN | 28.95 ± 0.42 | 28.12 [4] | 10K | 10K | netG.pth | dcgan_cifar.py | FID | WGAN-GP | 26.08 ± 0.12 | 29.3 [6] | 50K | 50K | netG.pth | wgangp32.py | FID | SNGAN | 23.90 ± 0.20 | 21.7 ± 0.21 [1]| 10K | 5K | netG.pth | sngan_32.py | FID | cGAN-PD | 17.84 ± 0.17 | 17.5 [2] | 10K | 5K | netG.pth | cganpd32.py | FID | SSGAN | 17.61 ± 0.14 | 17.88 ± 0.64 [3] | 10K | 10K | netG.pth | ssgan_32.py | FID | InfoMax-GAN | 17.14 ± 0.20 | 17.14 ± 0.20 [5] | 50K | 10K | netG.pth | infomaxgan32.py

Best FID was reported at 53K steps, but we find our score can improve till 100K steps to achieve 23.13 ± 0.13.

Citation

If you have found this work useful, please consider citing our work:

@article{lee2020mimicry,
    title={Mimicry: Towards the Reproducibility of GAN Research},
    author={Kwot Sin Lee and Christopher Town},
    booktitle={CVPR Workshop on AI for Content Creation},
    year={2020},
}

References

[1] Spectral Normalization for Generative Adversarial Networks

[2] cGANs with Projection Discriminator

[3] Self-Supervised GANs via Auxiliary Rotation Loss

[4] A Large-Scale Study on Regularization and Normalization in GANs

[5] InfoMax-GAN: Improved Adversarial Image Generation via Information Maximization and Contrastive Learning

[6] GANs Trained by a Two Time-Scale Update Rule Converge to a Local Nash Equilibrium

We use cookies. If you continue to browse the site, you agree to the use of cookies. For more information on our use of cookies please see our Privacy Policy.