Papers › Spectral Normalization for Generative Adversarial Networks

Spectral Normalization for Generative Adversarial Networks

16 Feb 2018ICLR 2018 1arXiv:1802.05957archive 2025-07-28

Takeru Miyato, Toshiki Kataoka, Masanori Koyama, Yuichi Yoshida

One of the challenges in the study of generative adversarial networks is the instability of its training. In this paper, we propose a novel weight normalization technique called spectral normalization to stabilize the training of the discriminator. Our new normalization technique is computationally light and easy to incorporate into existing implementations. We tested the efficacy of spectral normalization on CIFAR10, STL-10, and ILSVRC2012 dataset, and we experimentally confirmed that spectrally normalized GANs (SN-GANs) is capable of generating images of better or equal quality relative to the previous training stabilization techniques.

PaperPDFConference PDFCodeCode Syntology ran

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

For agents, Syntology's MCP tool lists every function and class Syntology harvested from this paper and whether it ran (how to connect): get_harvested_code_for_paper(arxiv_id="1802.05957")

Code

Syntology Ran 16 of 31 code samples harvested from 19 repositories linked to this paper; 15 have no recorded run. Of those that ran: 1 ran · violated contract; 2 ran · our draft was wrong; 1 ran · fixture could not drive it; 12 ran with no contract checked.

By repository: named in the paper: 1 sample from 1 repository, 1 ran; community (archive-listed): 30 samples from 18 repositories, 15 ran. The run record, sample by sample. “Ran” means executed on a synthesized input, not that the code is correct or reproduces the paper.

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

pfnet-research/sngan_projection officialmentioned in papertfNOASSERTION report
crcrpar/pytorch.sngan_projection mentioned in papermentioned on GitHubpytorch report
Bingwen-Hu/DRIT mentioned on GitHubpytorch report
Mohanned-Elkholy/ResNet-GAN mentioned on GitHubpytorch report
VitoRazor/Gan_Architecture mentioned on GitHubtfGPL-3.0 report
Xiaoming-Yu/SingleGAN mentioned on GitHubpytorchMIT report
adbobes/VideoSuperResolution mentioned on GitHubpytorch report
alhasan-abdellatif/cgans mentioned on GitHubpytorch report
apnkv/nla_spectral_norm mentioned on GitHubpytorch report
bayraktarbaris/SNGAN mentioned on GitHubpytorch report
guy-oren/DIRT-OST mentioned on GitHubpytorch report
hanyoseob/pytorch-sngan mentioned on GitHubpytorch report
hinofafa/Self-Attention-GAN mentioned on GitHubpytorch report
hinofafa/Self-Attention-HearthStone-GAN mentioned on GitHubpytorch report
jugatsingh/self_attention_video mentioned on GitHubpytorch report
karoly-hars/GAN_image_colorizing mentioned on GitHubpytorchMIT report
kiban/models mentioned on GitHub report
kklemon/bgan-pytorch mentioned on GitHubpytorch report
koshian2/SNGAN mentioned on GitHubpytorch report
ksagit/sn_gan_pytorch mentioned on GitHubpytorch report
meg965/pytorch-rl mentioned on GitHubpytorch report
ncuzzy/mygan mentioned on GitHubpytorchnot reachable when probed 2026-09-18 — repositories for recent papers often appear after camera-ready report
qiaolinhan/ws-preprocess mentioned on GitHub report
t-ae/singan-s4tf mentioned on GitHubtf report
yanqi1811/self-attention mentioned on GitHubpytorch report
zhusiling/SAGAN mentioned on GitHubpytorch report
zzmtsvv/adversarial mentioned on GitHubpytorch 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.

Code Syntology ran Syntology

31 samples harvested; 16 ran; 0 honoured the contract we drafted; 15 have no recorded run. Read from Syntology's graph 2026-09-24; that is when this build read the record, not when the samples ran.

1ran · violated contract
2ran · our draft was wrong
1ran · fixture could not drive it
12ran
15unverified

Licence: 14 of the 31 samples are pointer only, meaning Syntology does not serve that copy's text. This page shows no code text for any sample; each one links to its file in the repository.

Harvested from 19 repositories linked to this paper, official or community; each sample names its own and says which. “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. “Honoured” and “violated” refer to a contract Syntology drafted from the code itself; “our draft was wrong” and “fixture could not drive it” are failures of Syntology's instrument, not of the code.

Each sample ends with its code_sha256, Syntology's identity for that exact code. An agent fetches the stored sample with Syntology's MCP tool get_code(code_sha256="…") (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 from a repository marked official are listed first. 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 four 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; or the licence is outside the permissive list Syntology serves text under (MIT, Apache-2.0, BSD and similar). 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.

Block crcrpar/pytorch.sngan_projection/models/discriminators/resblocks.py named in the paper ran fingerprinted MIT (permissive) · 140b5bac7a8508a9 · report
DenseSN IShengFang/SpectralNormalizationKeras/SpectralNormalizationKeras.py community (archive-listed) ran licence not identified · pointer only · 1618dd723285892b · report
DenseSN GrayXu/SpectralNormalization-TF-Keras/SNorm_tf_keras.py community (archive-listed) ran MIT (permissive) · 78f554911984ea8d · report
ResidualBlock adbobes/VideoSuperResolution/models/frvsr.py community (archive-listed) ran · metamorphic tier: invariant no licence file found · pointer only · f202366742ae57fc · report
SNLinear ksagit/sn_gan_pytorch/spectral_layers.py community (archive-listed) ran no licence file found · pointer only · e5baa0b3b62b8c9b · report
SNLinear zzmtsvv/adversarial/sn_gan/sn_modules.py community (archive-listed) ran no licence file found · pointer only · 2ed2088d35626661 · report
SpectralNorm karolrogozinski/cern_alice_fast_sim_corrvae/utils/spectral_norm_fc.py community (archive-listed) ran no licence file found · pointer only · eee46d1790d3fc3d · report
SpectralNorm meg965/pytorch-rl/Layers/Spectral_norm.py community (archive-listed) ran · metamorphic tier: deterministic no licence file found · pointer only · c11fdae8c1eaa067 · report
SpectralNorm koshian2/SNGAN/models/core_layers.py community (archive-listed) ran · metamorphic tier: deterministic MIT (permissive) · a587e964e3b73e7f · report
SpectralNorm hanyoseob/pytorch-sngan/layer.py community (archive-listed) ran · metamorphic tier: deterministic no licence file found · pointer only · 95f6381171e5b22f · report
SpectralNorm apnkv/nla_spectral_norm/spec_norm_diff.py community (archive-listed) ran fingerprinted no licence file found · pointer only · eea0b4fc0256e64c · report
l2_normalization bayraktarbaris/SNGAN/spectral_normalization.py community (archive-listed) ran fingerprinted no licence file found · pointer only · 7a608922984c40b9 · report
l2_normalize zzmtsvv/adversarial/sn_gan/sn_modules.py community (archive-listed) ran · our draft was wrong fingerprinted no licence file found · pointer only · bf5a05c7adc49b0b · report
l2normalize christiancosgrove/pytorch-spectral-normalization-gan/spectral_normalization.py community (archive-listed) ran · our draft was wrong fingerprinted MIT (permissive) · bedff51745d2cf84 · report
l2normalize koshian2/SNGAN/models/core_layers.py community (archive-listed) ran · violated contract fingerprinted MIT (permissive) · 9b11458d7d1ec61c · report
max_singular_value ksagit/sn_gan_pytorch/spectral_layers.py community (archive-listed) ran · fixture could not drive it no licence file found · pointer only · 6ca4fd5871d2fad1 · report
KL_loss Xiaoming-Yu/SingleGAN/util/loss.py community (archive-listed) unverified MIT (permissive) · 64ebe64e3ec531c1 · report
ResBlockDiscriminator Mohanned-Elkholy/ResNet-GAN/src/discriminator/utils.py community (archive-listed) unverified MIT (permissive) · ad0eca31b5e574c4 · report
SN WangZesen/Spectral-Normalization-GAN/SNLayer.py community (archive-listed) unverified no licence file found · pointer only · 448cdd7f0cee0721 · report
SpectralNorm hinofafa/Self-Attention-HearthStone-GAN/spectral.py community (archive-listed) unverified no licence file found · pointer only · 5c6f3d4b49c16aa5 · report
SpectralNorm bayraktarbaris/SNGAN/spectral_normalization.py community (archive-listed) unverified no licence file found · pointer only · cb5d0b5f94066d27 · report
get_nl_layer Xiaoming-Yu/SingleGAN/models/model.py community (archive-listed) unverified MIT (permissive) · 146991926013af84 · report
get_norm_layer Xiaoming-Yu/SingleGAN/models/model.py community (archive-listed) unverified MIT (permissive) · 463156b57378ec6e · report
pickle_load Xiaoming-Yu/SingleGAN/util/util.py community (archive-listed) unverified MIT (permissive) · c4b9e0cb79abdf65 · report
postprocess karoly-hars/GAN_image_colorizing/datasets.py community (archive-listed) unverified MIT (permissive) · 128a565cc437a6cb · report
preprocess karoly-hars/GAN_image_colorizing/datasets.py community (archive-listed) unverified MIT (permissive) · 01d26fcf1d7994f7 · report
spectral_norm taki0112/Spectral_Normalization-Tensorflow/spectral_norm.py community (archive-listed) unverified MIT (permissive) · 4d8feb46b282de73 · report
tensor2im Xiaoming-Yu/SingleGAN/util/util.py community (archive-listed) unverified MIT (permissive) · 6c7988ea60c79ae4 · report
tensor2vec Xiaoming-Yu/SingleGAN/util/util.py community (archive-listed) unverified MIT (permissive) · 86091e3080eadd73 · report
unpickle_batch karoly-hars/GAN_image_colorizing/datasets.py community (archive-listed) unverified MIT (permissive) · a01c9b17a6119a4a · report
weights_init Xiaoming-Yu/SingleGAN/models/model.py community (archive-listed) unverified MIT (permissive) · 07d08fa9cc777740 · report

Tasks

Image Generation

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Image Generation CIFAR-10 SN-GANs FID 21.7 #58 of 78 Archive leaderboard report
Image Generation STL-10 SN-GAN FID 40.1 #26 of 31 Archive leaderboard report
Image Generation STL-10 SN-GAN Inception score 9.10 #26 of 31 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: SNGAN, Spectral Normalization

1x1 ConvolutionAverage PoolingBatch NormalizationBottleneck Residual BlockConvolutionDense ConnectionsGAN Hinge LossGlobal Average PoolingKaiming InitializationMax PoolingReLUResidual BlockResidual ConnectionSNGANSpectral NormalizationTanh Activation

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