Papers › Perceptual Adversarial Robustness: Defense Against Unseen Threat Models

Perceptual Adversarial Robustness: Defense Against Unseen Threat Models

22 Jun 2020arXiv:2006.12655archive 2025-07-28

Cassidy Laidlaw, Sahil Singla, Soheil Feizi

A key challenge in adversarial robustness is the lack of a precise mathematical characterization of human perception, used in the very definition of adversarial attacks that are imperceptible to human eyes. Most current attacks and defenses try to avoid this issue by considering restrictive adversarial threat models such as those bounded by L₂ or L_∞ distance, spatial perturbations, etc. However, models that are robust against any of these restrictive threat models are still fragile against other threat models. To resolve this issue, we propose adversarial training against the set of all imperceptible adversarial examples, approximated using deep neural networks. We call this threat model the neural perceptual threat model (NPTM); it includes adversarial examples with a bounded neural perceptual distance (a neural network-based approximation of the true perceptual distance) to natural images. Through an extensive perceptual study, we show that the neural perceptual distance correlates well with human judgements of perceptibility of adversarial examples, validating our threat model. Under the NPTM, we develop novel perceptual adversarial attacks and defenses. Because the NPTM is very broad, we find that Perceptual Adversarial Training (PAT) against a perceptual attack gives robustness against many other types of adversarial attacks. We test PAT on CIFAR-10 and ImageNet-100 against five diverse adversarial attacks. We find that PAT achieves state-of-the-art robustness against the union of these five attacks, more than doubling the accuracy over the next best model, without training against any of them. That is, PAT generalizes well to unforeseen perturbation types. This is vital in sensitive applications where a particular threat model cannot be assumed, and to the best of our knowledge, PAT is the first adversarial training defense with this property.

PaperPDFCodeCode Syntology ranMethod as code

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.

Syntology holds a reference implementation of Neural Perceptual Threat Model (NPTM), generated from the paper text (not the authors' code), at level 3 on Syntology's ladder: V2 “two implementations GENERATED from the paper's signature and spec agree numerically (1e-6) on fixture inputs” “neither of the two ever read the authors' own code, so V2 measures how unambiguously the paper specifies the method, not fidelity to the authors' implementation (measured 2026-09-11, commit 75eaa77: cross_checked_by is a second LLM on 2,628 of 2,831 served samples)”; V3 “paper-derived property tests pass on both of those same two generations.” (quoted from Syntology's server card). No rung compares this code against the authors' implementation. For agents: get_reference_implementation(name='Neural Perceptual Threat Model (NPTM)') (how to connect); the free check have('2006.12655') needs no token.

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

Code

Syntology 14 code samples harvested from 2 repositories linked to this paper: 11 ran (of which 9 constructed an object rather than computing a result; 10 with no instrument failure: 1 honoured, 0 violated, 9 with no contract checked; 1 where Syntology's instrument failed) · 3 unverified. Where Syntology's instrument failed: 1 where our contract draft was wrong; failures of Syntology's instrument, not of the code.

By repository: official repository: 12 samples from 1 repository, 9 ran (8 with no instrument failure; 1 where Syntology's instrument failed); community (archive-listed): 2 samples from 1 repository, 2 ran (2 with no instrument failure; 0 where Syntology's instrument failed). The run record, sample by sample. “Ran” means executed on a synthesized input, not that the code is correct or reproduces the paper.

Levels written on 2026-09-22 (UTC): the dates Syntology's graph wrote each sample's verification level, not when the code executed.

cassidylaidlaw/perceptual-advex officialmentioned in papermentioned on GitHubpytorchSyntology: 12 samples, 9 ran (1 instrument-failed) · 3 unverified report
centerforaisafety/adversarial-corruptions mentioned on GitHubpytorchSyntology: 2 samples, 2 ran (0 instrument-failed) · 0 unverified 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

14 samples harvested: 11 ran (of which 9 constructed an object rather than computing a result; 10 with no instrument failure: 1 honoured, 0 violated, 9 with no contract checked; 1 where Syntology's instrument failed) · 3 unverified. “Where Syntology's instrument failed” counts failures of Syntology's instrument, not of the code. Syntology's record for this page has not changed since , the first build that kept a record date for it; when this build read Syntology's graph is in the build record. That date is not when the samples ran.

1ran · honoured contract
1ran · our draft was wrong
9ran
3unverified

Licence: 2 of the 14 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 2 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. “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”. For 3 samples that ran, the graph records a signature this page does not show: for a class the fixture built without calling, the harness called the object with inputs it made up itself, so the recorded signature is the harness's probe, not the fixture's arguments. “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.

cassidylaidlaw/perceptual-advex official repository 12 samples, 9 ran (1 instrument-failed) · 3 unverified
AlexNetFeatureModel perceptual_advex/perceptual_attacks.py ran MIT (permissive) · 7145686c67025ec1 · report fixture: returned AlexNetFeatureModel · level recorded 2026-09-22
BisectionPerceptualProjection perceptual_advex/perceptual_attacks.py ran MIT (permissive) · 739afba6923ed65c · report fixture: returned BisectionPerceptualProjection · level recorded 2026-09-22
CifarAlexNet perceptual_advex/perceptual_attacks.py ran MIT (permissive) · e39bb656fc085ddb · report fixture: returned CifarAlexNet · level recorded 2026-09-22
ImageNetNormalizer perceptual_advex/perceptual_attacks.py ran fingerprinted MIT (permissive) · 74e9725f2c4515bd · report fixture: returned ImageNetNormalizer · level recorded 2026-09-22
MarginLoss perceptual_advex/perceptual_attacks.py ran MIT (permissive) · 5153783f8169f34c · report fixture: returned MarginLoss · level recorded 2026-09-22
NewtonsPerceptualProjection perceptual_advex/perceptual_attacks.py ran MIT (permissive) · 5b2c8693ad3f1ba0 · report fixture: returned NewtonsPerceptualProjection · level recorded 2026-09-22
NoProjection perceptual_advex/perceptual_attacks.py ran fingerprinted MIT (permissive) · 5569d8317369619e · report fixture: returned NoProjection · level recorded 2026-09-22
get_lpips_model perceptual_advex/perceptual_attacks.py ran · our draft was wrong MIT (permissive) · cf259b114c239eb5 · report fixture: signature lpips_model_spec:str/0/str, model:callable/0/? · returned MinimalAlexNetFeatureModel · level recorded 2026-09-22
normalize_flatten_features in 5 papers perceptual_advex/perceptual_attacks.py ran · honoured contract MIT (permissive) · 016b0a2f3d92b2bd · report fixture: signature features:tuple/1/Tensor, eps:float/0/float · returned Tensor · level recorded 2026-09-22
FastLagrangePerceptualAttack perceptual_advex/perceptual_attacks.py unverified MIT (permissive) · cc2c7f04f7f8cc5a · report
FeatureModel perceptual_advex/perceptual_attacks.py unverified MIT (permissive) · 3fd9de784fa31909 · report
LPIPSDistance perceptual_advex/perceptual_attacks.py unverified MIT (permissive) · af3b041cc30c1970 · report
centerforaisafety/adversarial-corruptions community (archive-listed) 2 samples, 2 ran (0 instrument-failed) · 0 unverified
Normalize in 154 papers models/cifar10/preactresnet18.py ran · metamorphic tier: deterministic fingerprinted no licence file found · pointer only · 5f40d79adb31f436 · report fixture: returned Normalize · level recorded 2026-09-22
PreActResNet models/cifar10/preactresnet18.py ran · metamorphic tier: deterministic no licence file found · pointer only · 86c37d42bbcb2386 · report fixture: returned PreActResNet · level recorded 2026-09-22

Method as code Syntology

What Syntology's graph records as this paper's claimed method, lifted into code: each assembly names its bricks (functions and classes), each with its own role and its own verification level. An assembly is not a check that the code implements the method correctly or reproduces the paper's results. Levels are each brick's own verification level in Syntology's graph (level 0: no recorded run), listed brick by brick and never combined into one level; “ran” means the brick executed on a synthesized input, not that its output is correct, and the instrument figure counts failures of Syntology's instrument, not of the code. For a generated assembly, an independent peer is a second implementation generated separately from the same paper text; neither implementation read the authors' code. At level 2 the two agreed numerically on the fixture's inputs; level 3 is recorded when property tests derived from the paper also passed on both, and each level 3 card prints how many its report lists. Where a model read the code against the paper's claim, each assembly's label verdict is one model's judgement, kept in this page's JSON twin; it is not calibrated, so no card prints it. A generated assembly's label was set when its code was generated for the method, and no model judged it; its card prints it as "by construction" or "not judged by a model". Siblings are assemblies lifted from this paper in the same batch for the same method: alternative readings kept side by side; none replaces or supersedes another. Where an assembly's phrase has no constructed figure, a brick that ran is not among the harvested samples this build reads, and whether a brick's run only constructed an object is read from that brick's harvested sample, so the figure was not counted; it does not say there were none.

3 assemblies on this paper: 2 harvested, 1 generated, listed in 2 groups (siblings lifted in one batch form one group).

generated Assembly for Neural Perceptual Threat Model (NPTM) (graph method: Neural Perceptual Threat Model (NPTM)), generated from the paper text, not the repository's code. 1 brick; individually on a fixture: 1 ran (1 with no instrument failure: 1 honoured, 0 violated, 0 with no contract checked; 0 where Syntology's instrument failed) · 0 unverified. Ran end-to-end against an independent peer (level 3; 5 property tests passed on both).

label matches by construction: the code was generated for this method (not judged by a model)

  1. multiply_jacobian entry · level 3 · ran · honoured contract · a91ae0c9df82f563ccdacd7b64486d7e588a41b0d0fcd23fc92ec0f6650585ad

2 sibling assemblies for Perceptual Adversarial Training (PAT) (harvested), lifted in the same batch on 2026-09-09: alternative readings kept side by side; none replaces or supersedes another.

harvested Assembly for Perceptual Adversarial Training (PAT) (graph method: Perceptual Adversarial Training (PAT)). 2 bricks; individually on a fixture: 2 ran (of which 2 constructed an object rather than computing a result; 2 with no instrument failure: 0 honoured, 0 violated, 2 with no contract checked; 0 where Syntology's instrument failed) · 0 unverified; every one of the 2 samples that ran constructed an object rather than computing a result. The assembly has not been run end-to-end.

one of 2 siblings

  1. PreActResNet entry · level 1 · ran · 86c37d42bbcb2386
  2. Normalize data_transform · level 1 · ran · 5f40d79adb31f436

harvested Assembly for Perceptual Adversarial Training (PAT) (graph method: Perceptual Adversarial Training (PAT)). 12 bricks; individually on a fixture: 9 ran (of which 7 constructed an object rather than computing a result; 8 with no instrument failure: 1 honoured, 0 violated, 7 with no contract checked; 1 where Syntology's instrument failed) · 3 unverified. The assembly has not been run end-to-end.

one of 2 siblings

  1. FastLagrangePerceptualAttack entry · level 0 · unverified · cc2c7f04f7f8cc5a
  2. LPIPSDistance helper · level 0 · unverified · af3b041cc30c1970
  3. MarginLoss helper · level 1 · ran · 5153783f8169f34c
  4. get_lpips_model model · level 1 · ran · our draft was wrong · cf259b114c239eb5
  5. normalize_flatten_features helper · level 1 · ran · honoured contract · 016b0a2f3d92b2bd
  6. AlexNetFeatureModel helper · level 1 · ran · 7145686c67025ec1
  7. FeatureModel helper · level 0 · unverified · 3fd9de784fa31909
  8. CifarAlexNet helper · level 1 · ran · e39bb656fc085ddb
  9. ImageNetNormalizer helper · level 1 · ran · 74e9725f2c4515bd
  10. BisectionPerceptualProjection helper · level 1 · ran · 739afba6923ed65c
  11. NewtonsPerceptualProjection helper · level 1 · ran · 5b2c8693ad3f1ba0
  12. NoProjection helper · level 1 · ran · 5569d8317369619e

Each brick ends with its code_sha256, Syntology's identity for that exact code; click one to copy the get_code call that returns it. For agents, on Syntology's MCP service: assembly(arxiv_id="2006.12655") returns these assemblies with each brick's role and wiring (how to connect).

Tasks

Adversarial DefenseAdversarial RobustnessPerceptual Distance

Results from the paper archive 2025-07-28

No leaderboard rows for this paper in the archive.

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