Papers › Fast AutoAugment

Fast AutoAugment

1 May 2019NeurIPS 2019 12arXiv:1905.00397archive 2025-07-28

Sungbin Lim, Ildoo Kim, Taesup Kim, Chiheon Kim, Sungwoong Kim

Data augmentation is an essential technique for improving generalization ability of deep learning models. Recently, AutoAugment has been proposed as an algorithm to automatically search for augmentation policies from a dataset and has significantly enhanced performances on many image recognition tasks. However, its search method requires thousands of GPU hours even for a relatively small dataset. In this paper, we propose an algorithm called Fast AutoAugment that finds effective augmentation policies via a more efficient search strategy based on density matching. In comparison to AutoAugment, the proposed algorithm speeds up the search time by orders of magnitude while achieves comparable performances on image recognition tasks with various models and datasets including CIFAR-10, CIFAR-100, SVHN, and ImageNet.

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="1905.00397")

Code

Syntology Ran 22 of 40 code samples harvested from 6 repositories linked to this paper; 18 have no recorded run. Of those that ran: 4 ran · our draft was wrong; 12 ran · fixture could not drive it; 6 ran with no contract checked.

By repository: official repository: 10 samples from 1 repository, 4 ran; community (archive-listed): 28 samples from 5 repositories, 17 ran; 2 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.

kakaobrain/fast-autoaugment officialmentioned in papermentioned on GitHubpytorchMIT report
Ching-Chen-Wang/EfficientNet-eLite mentioned on GitHubpytorch report
cfld/amdim mentioned on GitHubpytorch report
ildoonet/pytorch-randaugment mentioned on GitHubpytorch report
junkwhinger/fastautoaugment_jsh mentioned on GitHubpytorch report
kakaobrain/autoclint mentioned on GitHubpytorch report
philip-bachman/amdim-public mentioned on GitHubpytorchMIT report
songyadong106/111 mentioned on GitHubpytorchApache-2.0 report
tgilewicz/uniformaugment mentioned on GitHubpytorch report
zhanghang1989/fast-autoaug-torch mentioned on GitHubpytorchMIT report
NoamRosenberg/fast-autoaugment-on-a-single-gpu pytorchnot reachable when probed 2026-09-17 — repositories for recent papers often appear after camera-ready 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

40 samples harvested; 22 ran; 0 honoured the contract we drafted; 18 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.

4ran · our draft was wrong
12ran · fixture could not drive it
6ran
18unverified

Licence: 3 of the 40 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 6 repositories linked to this paper, official or community; each sample names its own and says which. Some samples are identical code Syntology first harvested from another repository; for those, this paper's copy is not located and its licence is not recorded. “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.

conv3x3 kakaobrain/fast-autoaugment/FastAutoAugment/networks/resnet.py official repository ran · our draft was wrong MIT (permissive) · fac5364e2f53c6db · report
conv3x3 kakaobrain/fast-autoaugment/FastAutoAugment/networks/pyramidnet.py official repository ran · our draft was wrong MIT (permissive) · 03a0be4fb7381bd2 · report
float_parameter kakaobrain/fast-autoaugment/archive.py official repository ran fingerprinted MIT (permissive) · 40d3c39a403ae1f7 · report
int_parameter kakaobrain/fast-autoaugment/archive.py official repository ran fingerprinted MIT (permissive) · 7900a3a6deef908b · report
accuracy kakaobrain/fast-autoaugment/FastAutoAugment/metrics.py official repository unverified MIT (permissive) · 3ccb35170e45f022 · report
get_logger kakaobrain/fast-autoaugment/FastAutoAugment/common.py official repository unverified MIT (permissive) · 6066adc7a58afde6 · report
mixup kakaobrain/fast-autoaugment/FastAutoAugment/aug_mixup.py official repository unverified MIT (permissive) · 8b2b8ba48892e73e · report
parse_devkit kakaobrain/fast-autoaugment/FastAutoAugment/imagenet.py official repository unverified MIT (permissive) · 10d196b211843339 · report
parse_meta kakaobrain/fast-autoaugment/FastAutoAugment/imagenet.py official repository unverified MIT (permissive) · 97bfaa2cce837d67 · report
parse_val_groundtruth kakaobrain/fast-autoaugment/FastAutoAugment/imagenet.py official repository unverified MIT (permissive) · e7825a1705689714 · report
Brightness tgilewicz/uniformaugment/UniformAugment/augmentations.py community (archive-listed) ran · fixture could not drive it MIT (permissive) · 993134d3c95bcf92 · report
Color tgilewicz/uniformaugment/UniformAugment/augmentations.py community (archive-listed) ran · fixture could not drive it MIT (permissive) · ccaa14b841bf44dc · report
Cutout tgilewicz/uniformaugment/UniformAugment/augmentations.py community (archive-listed) ran · fixture could not drive it MIT (permissive) · f90119cfdd38ef56 · report
Posterize tgilewicz/uniformaugment/UniformAugment/augmentations.py community (archive-listed) ran · fixture could not drive it MIT (permissive) · d2a9bd74240b6944 · report
Posterize ildoonet/pytorch-randaugment/RandAugment/augmentations.py community (archive-listed) ran · our draft was wrong MIT (permissive) · 75171b0fbf864dbb · report
Posterize2 tgilewicz/uniformaugment/UniformAugment/augmentations.py community (archive-listed) ran · fixture could not drive it MIT (permissive) · 067680c1cc9f4b97 · report
RandAugment ildoonet/pytorch-randaugment/RandAugment/augmentations.py community (archive-listed) ran MIT (permissive) · 9c6958943035b251 · report
ShearX kakaobrain/autoclint/skeleton/data/augmentations.py community (archive-listed) ran · fixture could not drive it Apache-2.0 (permissive) · b92936d1f2ac3e6c · report
ShearY kakaobrain/autoclint/skeleton/data/augmentations.py community (archive-listed) ran · fixture could not drive it Apache-2.0 (permissive) · 008c64b77c4ae4c9 · report
TranslateX kakaobrain/autoclint/skeleton/data/augmentations.py community (archive-listed) ran · fixture could not drive it Apache-2.0 (permissive) · ab50d5bbcf589623 · report
TranslateXAbs tgilewicz/uniformaugment/UniformAugment/augmentations.py community (archive-listed) ran · fixture could not drive it MIT (permissive) · b1fc55be13be06e7 · report
TranslateY tgilewicz/uniformaugment/UniformAugment/augmentations.py community (archive-listed) ran · fixture could not drive it MIT (permissive) · d4b73cc9fa9b08ae · report
TranslateYAbs tgilewicz/uniformaugment/UniformAugment/augmentations.py community (archive-listed) ran · fixture could not drive it MIT (permissive) · 15ceb820d17f0a5c · report
UniformAugment tgilewicz/uniformaugment/UniformAugment/augmentations.py community (archive-listed) ran MIT (permissive) · 1ca02113ff95139e · report
augment_list tgilewicz/uniformaugment/UniformAugment/augmentations.py community (archive-listed) ran · our draft was wrong fingerprinted MIT (permissive) · b1cd4e4b4bba7f8c · report
augment_list ildoonet/pytorch-randaugment/RandAugment/augmentations.py community (archive-listed) ran MIT (permissive) · 5d568852c0fdc6a3 · report
fa_reduced_cifar10 Ching-Chen-Wang/EfficientNet-eLite/EfficientNet-eLite/archive.py community (archive-listed) ran · metamorphic tier: deterministic no licence file found · pointer only · a43ccb0a81208bab · report
SolarizeAdd ildoonet/pytorch-randaugment/RandAugment/augmentations.py community (archive-listed) unverified MIT (permissive) · fcce9b35a12cd32c · report
count_bn2d songyadong106/111/skeleton/nn/modules/profile.py community (archive-listed) unverified Apache-2.0 (permissive) · 0a6d90cf17a6a7b6 · report
count_conv2d songyadong106/111/skeleton/nn/modules/profile.py community (archive-listed) unverified Apache-2.0 (permissive) · 2495aa8db734b762 · report
count_relu songyadong106/111/skeleton/nn/modules/profile.py community (archive-listed) unverified Apache-2.0 (permissive) · 8adf833ce8491ae7 · report
decorator_tuple_to_args songyadong106/111/skeleton/nn/modules/wrappers.py community (archive-listed) unverified Apache-2.0 (permissive) · 5d3725d10f387239 · report
get_change_scale songyadong106/111/skeleton/optim/scheduler.py community (archive-listed) unverified Apache-2.0 (permissive) · 500151731fe0c778 · report
get_discrete_epoch songyadong106/111/skeleton/optim/scheduler.py community (archive-listed) unverified Apache-2.0 (permissive) · fd8ab66e1567e6f2 · report
get_logger songyadong106/111/skeleton/projects/others.py community (archive-listed) unverified Apache-2.0 (permissive) · 58895526cd27ed71 · report
get_tf_resize songyadong106/111/skeleton/projects/others.py community (archive-listed) unverified Apache-2.0 (permissive) · 305d5112606704d9 · report
get_tf_to_tensor songyadong106/111/skeleton/projects/others.py community (archive-listed) unverified Apache-2.0 (permissive) · 58be042029debe2d · report
gradual_warm_up songyadong106/111/skeleton/optim/scheduler.py community (archive-listed) unverified Apache-2.0 (permissive) · 5f37303469d04cab · report
split_dataset_train identical code first harvested elsewhere ran · fixture could not drive it licence of this copy not recorded · f8a38db0674e2431 · report
generate_cv_sets identical code first harvested elsewhere unverified licence of this copy not recorded · 6f457a08830581a5 · report

Tasks

Data AugmentationImage AugmentationImage Classification

1 archive task tag without a task page not shown.

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Data Augmentation ImageNet ResNet-200 (Fast AA) Accuracy (%) 80.6 #4 of 17 Archive leaderboard report
Data Augmentation ImageNet ResNet-50 (Fast AA) Accuracy (%) 77.6 #13 of 17 Archive leaderboard report
Image Classification CIFAR-10 PyramidNet+ShakeDrop (Fast AA) Percentage correct 98.3 #46 of 265 Archive leaderboard report
Image Classification CIFAR-100 PyramidNet+ShakeDrop (Fast AA) Percentage correct 88.3 #38 of 211 Archive leaderboard report
Image Classification ImageNet ResNet-200 (Fast AA) Top 1 Accuracy 80.6% #691 of 1060 Archive leaderboard report
Image Classification ImageNet ResNet-50 (Fast AA) Top 1 Accuracy 77.6% #867 of 1060 Archive leaderboard report
Image Classification SVHN Wide-ResNet-28-10 (Fast AA) Percentage error 1.1 #2 of 62 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: Fast AutoAugment

1x1 ConvolutionAutoAugmentAverage PoolingBatch NormalizationBottleneck Residual BlockColorJitterConvolutionCutoutDropoutFast AutoAugmentGlobal Average PoolingKaiming InitializationLSTMMax PoolingReLUResidual BlockResidual ConnectionSigmoid ActivationTanh ActivationWeight DecayWide Residual BlockWideResNet

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