Papers › Training Generative Adversarial Networks with Limited Data

Training Generative Adversarial Networks with Limited Data

11 Jun 2020NeurIPS 2020 12arXiv:2006.06676archive 2025-07-28

Tero Karras, Miika Aittala, Janne Hellsten, Samuli Laine, Jaakko Lehtinen, Timo Aila

Training generative adversarial networks (GAN) using too little data typically leads to discriminator overfitting, causing training to diverge. We propose an adaptive discriminator augmentation mechanism that significantly stabilizes training in limited data regimes. The approach does not require changes to loss functions or network architectures, and is applicable both when training from scratch and when fine-tuning an existing GAN on another dataset. We demonstrate, on several datasets, that good results are now possible using only a few thousand training images, often matching StyleGAN2 results with an order of magnitude fewer images. We expect this to open up new application domains for GANs. We also find that the widely used CIFAR-10 is, in fact, a limited data benchmark, and improve the record FID from 5.59 to 2.42.

PaperPDFConference PDFCodeCode Syntology ran

In Syntology View this paper on Syntology, its page in Syntology's graph. That page adds which of its functions are also present in other papers' code Syntology holds, with dates, and its dated run record as JSON and a README badge. It also lists the repositories linked to the paper, the abstract and the calls for agents.

Open this paper in Syntology's Atlas, the map of the papers in Syntology's graph and their citations.

For agents, on Syntology's MCP service (how to connect):

Code

The archive's official repository (NVlabs/stylegan2-ada) was not harvested; the 28 samples below come from 5 community repositories, and 1 more is identical to code Syntology harvested from another paper's repository.

Syntology 29 code samples: 28 from 5 repositories linked to this paper, 1 identical to code harvested from another paper's repository. Across all 29: 21 ran (of which 0 constructed an object rather than computing a result; 19 with no instrument failure: 2 honoured, 1 violated, 16 with no contract checked; 2 where Syntology's instrument failed) · 8 unverified. Where Syntology's instrument failed: 1 where our contract draft was wrong; 1 where the contract check could not run; failures of Syntology's instrument, not of the code.

By repository: community (archive-listed): 28 samples from 5 repositories, 20 ran (18 with no instrument failure; 2 where Syntology's instrument failed); 1 identical to code first harvested elsewhere. The run record, sample by sample. “Ran” means executed on a synthesized input, not that the code is correct or reproduces the paper.

No sample is linked to this paper from 3 of its repositories that Syntology harvested for other papers:

Levels written between 2026-09-22 and 2026-09-26 (UTC): the dates Syntology's graph wrote each sample's verification level, not when the code executed. 1 of the 21 samples that ran has no recorded date.

28 repositories listed; official and paper-mentioned ones first.

NVlabs/stylegan2-ada officialmentioned in papermentioned on GitHubtfNOASSERTIONSyntology: not harvested report
BearNinja123/StyleGAN_ADAnough mentioned on GitHubtfSyntology: not harvested report
Eitan177/testGenImages mentioned on GitHubpytorchNOASSERTIONSyntology: not harvested report
NVlabs/stylegan2-ada-pytorch mentioned on GitHubpytorchSyntology: 1 sample, 1 ran (0 instrument-failed) · 0 unverified report
NariMo91/GANs-generative-art mentioned on GitHubpytorchSyntology: not harvested report
aiksir/stylegan2-ada-blending mentioned on GitHubtfNOASSERTIONSyntology: not harvested report
beresandras/gan-flavours-keras mentioned on GitHubtfMITSyntology: 4 samples, 1 ran (0 instrument-failed) · 3 unverified report
buganart/stylegan2-ada-pytorch mentioned on GitHubpytorchNOASSERTIONSyntology: not harvested report
datduong/stylegan2-ada-Ws-22q mentioned on GitHubtfNOASSERTIONSyntology: not harvested report
duskvirkus/stylegan2-ada-lightning mentioned on GitHubpytorchNOASSERTIONSyntology: not harvested report
eps696/stylegan2ada mentioned on GitHubpytorchNOASSERTIONSyntology: not harvested report
esimpsontheartist/stylegan2-FineArt mentioned on GitHubtfNOASSERTIONSyntology: not harvested report
fai07600521/final-project mentioned on GitHubpytorchNOASSERTIONSyntology: not harvested report
jiangshuyi0v0/cvd-gan mentioned on GitHubpytorchSyntology: not harvested report
lelechen63/stylegannerf mentioned on GitHubpytorchNOASSERTIONSyntology: not harvested report
mahmoudnafifi/HistoGAN mentioned on GitHubpytorchSyntology: no sample linked to this paper (harvested for another paper). Syntology's graph links these samples to HistoGAN: Controlling Colors of GAN-Generated and Real Images via Color Histograms (arXiv:2011.11731; that paper's own run record: 2 ran (of which 0 constructed an object rather than computing a result; 1 with no instrument failure: 1 honoured, 0 violated, 0 with no contract checked; 1 where Syntology's instrument failed) · 0 unverified); nothing checks that they implement this paper's method. report
matjazmav/fri-2021-ibb-seminar mentioned on GitHubpytorchSyntology: not harvested report
pbaylies/stylegan2 mentioned on GitHubtfNOASSERTIONSyntology: not harvested report
sangyun884/Face2Webtoon mentioned on GitHubpytorchMITSyntology: 14 samples, 9 ran (1 instrument-failed) · 5 unverified report
sh4174/3d-stylegan2-ada mentioned on GitHubtfNOASSERTIONSyntology: not harvested report
usufyan29/stylegan2_runway mentioned on GitHubtfNOASSERTIONSyntology: not harvested report
vsemecky/stylegan2-ada mentioned on GitHubtfNOASSERTIONSyntology: not harvested report
wangamelia/cmpm202p2p2 mentioned on GitHubtfSyntology: 2 samples, 2 ran (1 instrument-failed) · 0 unverified report
wjdals3406/stylegan2-ada-encoder mentioned on GitHubtfNOASSERTIONSyntology: not harvested report
woctezuma/steam-stylegan2-ada mentioned on GitHubpytorchMITSyntology: not harvested report
ChristophReich1996/Multi-StyleGAN pytorchMITSyntology: 7 samples, 7 ran (0 instrument-failed) · 0 unverified report
keras-team/keras-io tfSyntology: no sample linked to this paper (harvested for another paper). Syntology's graph links these samples to The Forward-Forward Algorithm: Some Preliminary Investigations (arXiv:2212.13345; that paper's own run record: 18 ran (of which 0 constructed an object rather than computing a result; 11 with no instrument failure: 0 honoured, 2 violated, 9 with no contract checked; 7 where Syntology's instrument failed) · 2 unverified); nothing checks that they implement this paper's method. report
toshas/torch-fidelity pytorchSyntology: no sample linked to this paper (harvested for another paper). Syntology's graph links these samples to Improved Precision and Recall Metric for Assessing Generative Models (arXiv:1904.06991; that paper's own run record: 30 ran (of which 12 constructed an object rather than computing a result; 20 with no instrument failure: 5 honoured, 1 violated, 14 with no contract checked; 10 where Syntology's instrument failed) · 12 unverified); nothing checks that they implement this paper's method. report

Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.

A paper named beside a repository with no sample linked to this paper is shown with that paper's own run record, not this paper's: “ran” means executed on a synthesized input, not that the code is correct, and “where Syntology's instrument failed” counts failures of Syntology's instrument, not of the code. Papers are listed in arXiv-id order, at most three per repository.

Code Syntology ran Syntology

29 samples harvested: 21 ran (of which 0 constructed an object rather than computing a result; 19 with no instrument failure: 2 honoured, 1 violated, 16 with no contract checked; 2 where Syntology's instrument failed) · 8 unverified. “Where Syntology's instrument failed” counts failures of Syntology's instrument, not of the code. Syntology's record for this page last changed ; when this build read Syntology's graph is in the build record. That date is not when the samples ran.

2ran · honoured contract
1ran · violated contract
1ran · our draft was wrong
1ran; the contract check could not run
16ran
8unverified

Licence: 5 of the 29 samples are pointer only, meaning Syntology does not serve that copy's text. For 1 of them, this paper's copy is under a permissive licence and cleared, but the stored copy of the same code that Syntology serves is not; the licence label's tooltip gives Syntology's recorded reason. This page shows no code text for any sample; each one links to its file in the repository.

Harvested from 5 repositories linked to this paper, official or community; the list below is grouped by repository, each group headed by its name and what it is to the paper. 1 sample is listed last, in a group without a repository. “Ran” means the sample executed on a synthesized input. It does not mean the output is correct, and nothing here reproduces the paper's results. A line beginning “fixture:” under a sample that ran gives the arguments Syntology's fixture called the sample with, each as name:type/rank/element kind (? where none was recorded), and the type of what the fixture's run returned, exactly as the harness recorded them; the arguments and the returned type were recorded by two separate runs of the same fixture. No output value is shown, and a returned type says nothing about whether the output was right. “Level recorded” gives the UTC date when Syntology's graph wrote the sample's verification level, not when the code executed; where the level was written before Syntology's graph kept that date, the row says “level recorded, date not recorded”. “Honoured” and “violated” refer to a contract Syntology drafted from the code itself. “Our draft was wrong” and “the contract check could not run” are failures of Syntology's instrument, not of the code: the sample ran on Syntology's synthesized input, and then the drafted contract did not describe it, or the check itself raised an error (RAISES), lacked a dependency (DEP_MISSING) or ran out of time (TIMEOUT); hover the label for which. A metamorphic tier, where shown, is evidence of the output's shape, not that the sample implements the method.

Each sample ends with its code_sha256. A sample's code_sha256 is Syntology's identity for that exact code. It is printed in full (16 hexadecimal characters for every sample in this build), and get_code(code_sha256="…") takes it exactly as printed. An agent calls it on Syntology's MCP service (how to connect); click an identity to copy that call.

Repository labels, per sample. official repository: The archive marks this repository official for the paper. named in the paper: The archive records that the paper mentions this repository; it is not marked official. community (archive-listed): In the archive's code links for this paper, not marked official and not recorded as mentioned in the paper. found in paper text by Syntology: Syntology found this repository in the paper's own text; whether it is the authors' implementation is not asserted. community: Not in the archive's code links for this paper; a community repository Syntology harvested. Samples are grouped by repository: a repository marked official first, then by how directly the repository is tied to the paper, then in the archive's order; never by how many ran. Licence labels name the repository's licence as recorded at harvest. “Pointer only” means Syntology does not serve that copy's text, for one of six reasons: no licence file was found; the licence was not identified; the licence is recorded as permissive but that copy's record is not marked cleared; the licence is outside the permissive list Syntology serves text under (MIT, Apache-2.0, BSD and similar); this paper's copy was not located, so its licence is not recorded; or this paper's copy is cleared, but the stored copy of the same code that Syntology serves is not. Some licences outside that list permit redistribution, such as WTFPL, and GPL-3.0 under its conditions; they are simply not on the list. Hover a licence label for the reason. File links open the file on GitHub at the default branch, which may have changed since the harvest.

A function or class name links to its page when the name occurs in the harvested code of at least 5 papers; the name is a string, not an identity. A census tag appears when Syntology's census ran implementations with that name on one shared input per bucket: it gives how many were compared, from how many papers, and how many distinct outputs they produced, and names this sample's bucket when this sample was one of them.

NVlabs/stylegan2-ada-pytorch community (archive-listed) 1 sample, 1 ran (0 instrument-failed) · 0 unverified
num_range in 10 papers style_mixing.py ran · honoured contract licence not identified · pointer only · 5c53150c611eaf49 · report fixture: signature s:str/0/str · returned list · level recorded 2026-09-22
beresandras/gan-flavours-keras community (archive-listed) 4 samples, 1 ran (0 instrument-failed) · 3 unverified
preprocess_birds dataset.py ran MIT (permissive) · ed196116d992670e · report fixture: signature image_size:int/0/int, padding:float/0/float · returned function · level recorded 2026-09-24
preprocess_flowers dataset.py unverified MIT (permissive) · f531ed61af7fc0f2 · report
round_to_int dataset.py unverified MIT (permissive) · f7831e90f18bf2e0 · report
step in 32 papers utils.py unverified MIT (permissive) · e4aa47838d197a6b · report
sangyun884/Face2Webtoon community (archive-listed) 14 samples, 9 ran (1 instrument-failed) · 5 unverified
calculate_activation_statistics in 75 papers gan_metrics_pytorch/fid_score.py ran fingerprinted MIT (permissive) · fa02bf7e7932504d · report fixture: signature act:ndarray/2/float · returned tuple · level recorded 2026-09-24
extract_lenet_features gan_metrics_pytorch/kid_score.py ran MIT (permissive) · ad1334727ae62246 · report fixture: signature imgs:ndarray/4/float, net:callable/0/? · returned ndarray · level recorded 2026-09-24
find_classes in 61 papers dataset.py ran · our draft was wrong MIT (permissive) · 52d10a10fb72fb9f · report fixture: signature dir:str/0/str · returned tuple · level recorded 2026-09-24
has_file_allowed_extension in 38 papers dataset.py ran MIT (permissive) · e800290c5ea55efd · report fixture: signature filename:str/0/str, extensions:list/1/str · returned bool · level recorded 2026-09-26
integer_translation ADA.py ran MIT (permissive) · 2296f6890f1691ee · report fixture: signature img:Image/0/?, p:float/0/float, r:float/0/float · returned Image · level recorded 2026-09-24
make_dataset in 144 papers dataset.py ran MIT (permissive) · 273d5df61b8a0cd5 · report fixture: signature dir:str/0/str, extensions:tuple/1/str, target:int/0/int · returned list · level recorded 2026-09-24
preprocessing in 22 papers utils.py ran fingerprinted MIT (permissive) · eab7ba887c245523 · report fixture: signature x:Tensor/2/float · returned Tensor · level recorded 2026-09-26
random_rotation90 ADA.py ran MIT (permissive) · 5377fb88325fa9b5 · report fixture: signature img:Image/0/?, p:float/0/float · returned Image · level recorded 2026-09-24
x_flip ADA.py ran MIT (permissive) · 56f419ded8d6d63a · report fixture: signature img:Image/0/?, p:float/0/float · returned Image · level recorded 2026-09-24
calculate_frechet_distance in 119 papers gan_metrics_pytorch/fid_score.py unverified MIT (permissive) · 0def50a351111624 · report
get_activations in 100 papers gan_metrics_pytorch/fid_score.py unverified MIT (permissive) here; the stored copy is pointer only · 7b41327814ab3634 · report
get_activations in 100 papers gan_metrics_pytorch/kid_score.py unverified MIT (permissive) · 0bdaf2a87d584365 · report
load_test_data in 21 papers utils.py unverified MIT (permissive) · 8d4ee7233484c7ee · report
save_images in 24 papers utils.py unverified MIT (permissive) · 5b8b7a85da8c0a00 · report
wangamelia/cmpm202p2p2 community (archive-listed) 2 samples, 2 ran (1 instrument-failed) · 0 unverified
create_image_grid in 6 papers generate.py ran; the contract check could not run (RAISES) · metamorphic tier: invariant licence not identified · pointer only · d2790eb5fa6571da · report fixture: signature images:ndarray/4/float, grid_size:tuple/1/int · returned ndarray · level recorded 2026-09-22
valmap generate.py ran · violated contract licence not identified · pointer only · c5e25d2754ccb9a0 · report fixture: signature value:float/0/float, istart:float/0/float, istop:float/0/float, ostart:float/0/float, ostop:float/0/float · returned float · level recorded 2026-09-22
ChristophReich1996/Multi-StyleGAN community (archive-listed) 7 samples, 7 ran (0 instrument-failed) · 0 unverified
generate_cut_mix_augmentation_data multi_stylegan/u_net_2d_discriminator.py ran fingerprinted MIT (permissive) · dfbcba5442b1b006 · report fixture: signature image_real:Tensor/4/float, image_fake:Tensor/4/float · returned tuple · level recorded 2026-09-24
generate_cut_mix_transformation_data multi_stylegan/u_net_2d_discriminator.py ran MIT (permissive) · 4b65e212777706b6 · report fixture: signature image_real:Tensor/4/float, image_fake:Tensor/4/float, prediction_real:Tensor/4/float, prediction_fake:Tensor/4/float · returned tuple · level recorded 2026-09-24
integer_translation multi_stylegan/adaptive_discriminator_augmentation.py ran fingerprinted MIT (permissive) · 3ecf4f91a6ada17a · report fixture: signature images:Tensor/4/float · returned Tensor · level recorded 2026-09-24
normalize_0_1_batch multi_stylegan/misc.py ran fingerprinted MIT (permissive) · d1307ae3434487a3 · report fixture: signature input:Tensor/4/float · returned Tensor · level recorded 2026-09-24
normalize_m1_1_batch multi_stylegan/misc.py ran fingerprinted MIT (permissive) · fc0c300a2ed629cf · report fixture: signature input:Tensor/5/float · returned Tensor · level recorded 2026-09-24
random_permutation multi_stylegan/misc.py ran MIT (permissive) · 1ce98174858bf663 · report fixture: signature n:int/0/int · returned Tensor · level recorded 2026-09-24
upfirdn2d_native in 26 papers
census: 5 implementations from 7 papers, 1 distinct output on one shared input
multi_stylegan/op_static/upfirdn2d.py ran MIT (permissive) · 6ff3df37d07d4cf5 · report fixture: signature input:Tensor/4/float, kernel:Tensor/2/float, up_x:int/0/int, up_y:int/0/int, down_x:int/0/int, down_y:int/0/int, pad_x0:int/0/int, pad_x1:int/0/int, pad_y0:int/0/int, pad_y1:int/0/int · returned Tensor · level recorded 2026-09-24
Repository not located 1 sample, 1 ran (0 instrument-failed) · 0 unverified
1 sample is identical to code Syntology harvested from another paper's repository; this paper's own copy was not located, so no repository is named and no licence is recorded.
num_range in 10 papers repository not located ran · honoured contract licence of this copy not recorded · 6f6beb9107cb48e8 · report fixture: signature s:str/0/str · returned list · level recorded, date not recorded

Tasks

10-shot image generationConditional Image GenerationImage Generation

Datasets

Introduced by this paper, per the archive.

MetFaces

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
10-shot image generation Babies TGAN + ADA FID 97.91 #6 of 7 Archive leaderboard report
Conditional Image Generation ArtBench-10 (32x32) StyleGAN2 + ADA FID 2.625 #1 of 6 Archive leaderboard report
Conditional Image Generation CIFAR-10 StyleGAN2-ADA FID 2.42 #4 of 25 Archive leaderboard report
Conditional Image Generation CIFAR-10 StyleGAN2-ADA Inception score 10.14 #4 of 25 Archive leaderboard report
Image Generation AFHQ Cat StyleGAN2-ADA FID 3.55 #2 of 8 Archive leaderboard report
Image Generation AFHQ Cat StyleGAN2-ADA clean-FID 3.28 ± .02 #2 of 8 Archive leaderboard report
Image Generation AFHQ Cat StyleGAN2-ADA clean-KID 0.71 ± .02 #2 of 8 Archive leaderboard report
Image Generation AFHQ Dog StyleGAN2-ADA FID 7.41 #4 of 6 Archive leaderboard report
Image Generation AFHQ Dog StyleGAN2-ADA clean-FID 7.61 ± .02 #4 of 6 Archive leaderboard report
Image Generation AFHQ Dog StyleGAN2-ADA clean-KID 1.28 ± .02 #4 of 6 Archive leaderboard report
Image Generation AFHQ Wild StyleGAN2-ADA FID 3.05 #2 of 5 Archive leaderboard report
Image Generation AFHQ Wild StyleGAN2-ADA clean-FID 3.00 ± .01 #2 of 5 Archive leaderboard report
Image Generation AFHQ Wild StyleGAN2-ADA clean-KID 0.44 ± .01 #2 of 5 Archive leaderboard report
Image Generation FFHQ 1024 x 1024 StyleGAN2 ADA+bCR FID 3.62 #10 of 20 Archive leaderboard report
Image Generation FFHQ 256 x 256 StyleGAN2 + ADA FID 3.62 #20 of 51 Archive leaderboard report
Image Generation FFHQ 256 x 256 StyleGAN2 + ADA (DINOv2) FD 514.78 #48 of 51 Archive leaderboard report
Image Generation FFHQ 256 x 256 StyleGAN2 + ADA (DINOv2) Precision 0.59 #48 of 51 Archive leaderboard report
Image Generation FFHQ 256 x 256 StyleGAN2 + ADA (DINOv2) Recall 0.06 #48 of 51 Archive leaderboard report
Image Generation Pokemon 256x256 StyleGAN2-ADA FID 40.38 #3 of 4 Archive leaderboard report

Ranks are positions in the archive's leaderboards as they stood at the 2025-07-28 snapshot. Results published since then are not among these rows, so a rank here is not a current standing.

Methods

Introduced by this paper: ADA

1x1 ConvolutionADAAdamBatch NormalizationBigGANConditional Batch NormalizationConvolutionDense ConnectionsEarly StoppingFeedforward NetworkGAN Hinge LossLinear LayerNon-Local BlockNon-Local OperationOff-Diagonal Orthogonal RegularizationPath Length RegularizationProjection DiscriminatorR1 RegularizationRandAugmentReLUResidual BlockResidual ConnectionSAGANSoftmaxSpectral NormalizationStyleGAN2Truncation TrickWeight Demodulation

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections