Now You See the Hate: Adaptive View Retrieval for Hidden Hateful Illusions

Derya Soydaner; Qianpu Chen

REVIEW 3 major objections 4 minor 33 references

Now You See the Hate: Adaptive View Retrieval for Hidden Hateful Illusions

T0 review · 3 major / 4 minor · reviewed 2026-08-01 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read Hate hidden inside optical illusions can be reliably detected by adaptively retrieving the concealed message from a bank of known templates, rather than by classifying the original view.

desk verdict Closed-set result is solid and well-analyzed, but the paper never quantifies the admitted drop on unseen symbol families, which is the number that matters. read the letter →

arxiv 2607.19061 v2 pith:NAKTUR42 submitted 2026-07-21 cs.CV cs.AI

Qianpu Chen , Derya Soydaner This is my paper · ORCID

classification cs.CVcs.AI

keywords hatefulillusionsmultimodalmoderationvisualretrievaladaptiveviewselectiontemplatebankCLIPhiddencontentoptical

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper tries to establish that hidden hateful content in optical illusions is primarily an observation problem, not a classification problem: moderation systems fail because they never see the concealed message. Its proposed solution, Adaptive View Retrieval, applies the same set of seven perceptual transforms (contrast, edge closure, figure-ground separation, low-pass smoothing, plus the original) to both the query image and a bank of known message templates, lets a learned gate choose which view each image should be matched in, retrieves the best-matching template, and then calibrates a separate harmful/benign decision. On the HatefulIllusion benchmark with a frozen CLIP encoder, this reaches 93.2% balanced accuracy on the held-out test split, against roughly chance-level performance for original-view classifiers and VLMs. A sympathetic reader would care because the same adaptive view bank also transfers to hidden-digit, hidden-object, and hidden-text recognition on other benchmarks, suggesting a general recipe for recovering hidden meaning before moderation.

What carries the argument

The load-bearing mechanism is Adaptive View Retrieval, a retrieve-and-calibrate architecture. Its core object is a complementary view bank: seven deterministic image operators (original, contrast-enhanced, closure-edge, three figure-ground variants, and low-pass) applied identically to the query and every template, so retrieval compares matched views in a shared embedding space. A frozen CLIP encoder produces view embeddings, branch MLPs project them, and a learned gate reallocates trust from view to view depending on target type and visibility. Template scoring is a temperature-scaled, gated sum of cosine similarities, and a separate calibration MLP reads the gated fusion to produce the har

What would settle it

Held-out discovery test: take the HatefulIllusion test set, remove all templates belonging to one hate-symbol family (e.g., all variants of one symbol) from the bank, and measure balanced accuracy on those images. If accuracy collapses or those images get mislabeled as a different known template, the closed-set dependence is confirmed; if the system flags them as uncertain or unknown, the retrieval can in principle support open-set moderation. A second observation: report the exact 'stricter target holdout' number the authors acknowledge but do not provide.

Watch

Extended reading notes

Core claim

The paper's central claim is that hidden hateful illusions can be detected by reframing detection as perceptual retrieval. Instead of a K-way label head on the original image, the authors maintain a hidden-message template bank and retrieve the message identity through matched-view comparisons: the same set of seven classical perceptual views is generated for the input and for every template, a learned gate weights the views per image, and a frozen CLIP encoder embeds each view so retrieval can compare aligned evidence. The retrieved identity then feeds a calibration head that decides whether the evidence is harmful. The paper reports 93.2% balanced accuracy (mean ±3.1 over three seeds) on t

Load-bearing premise

The load-bearing premise is that every hidden message at test time has a reference template in the bank: retrieval picks the best-scoring known template, so an unseen hate string or an unseen symbol family is at best labeled as its closest known neighbor, and the paper explicitly notes performance drops under stricter target holdout on unseen symbol families.

Editorial extensions

If this is right

Platforms that maintain template libraries of known slurs and symbols could add new threats without retraining the detector, and could audit a decision by inspecting which view and which template drove it.
No single classical transform is a sufficient safety filter; only a complementary bank with learned gating keeps performance stable on hard low-visibility and benign look-alike cases.
Retrieval-only training collapses moderation to 47.8%, so any practical deployment needs joint identity-retrieval and harm-calibration training.
The same view bank transfers to non-hate hidden content (digits, fashion, animals, text), so the mechanism is a general perceptual-recovery module rather than a hate-specific classifier.
On HC-Bench, multi-view questioning recovers hidden text and objects where resolution-losing zoom-out fails (98.2% vs 2.7%), suggesting adaptive multi-view input can replace or supplement downsampling in VLM pipelines.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

Editorial inference: Because retrieval is argmax over a fixed bank, the method's practical ceiling is template coverage; a natural extension is a rejection or 'unknown message' option based on retrieval confidence or a threshold on the top similarity score, which the paper does not implement.
Editorial inference: The gate's visibility-dependent reallocation (more low-pass weight when signals are faint) suggests visibility labels could be used as weak supervision or as a conditioning variable, potentially improving low-visibility recall further.
Editorial inference: The same matched-view bank could be tested on adversarial evasions—images crafted specifically to steer the gate toward a misleading view—since the failure analysis already shows deterministic figure-ground segmentation can break on textured carriers.
Editorial inference: The HC-Bench result with a small 2B VLM raises a testable hypothesis: multiplicative gains may come from giving VLMs multiple observation strategies rather than trading resolution for zoom; a direct comparison on a much larger VLM would clarify whether the view bank is a substitute for scale or only a fallback when scale is unavailable.

Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

Short version: the headline number is real for what it measures, but it measures a closed-set retrieval task. Every test message has a template in M, so 93.2% tells you the method can retrieve known messages if you point it at the right candidates. The authors admit in their Limitations section that performance drops on unseen symbol families and that this is 'not open-vocabulary discovery of unseen hate strings'—but they never report the magnitude. That's the number that matters for the stated motivation, and it's missing.

What's genuinely new: the combination of adaptive gating over a deterministic perceptual view bank with CLIP template retrieval and a separate calibration head. Prior work on these benchmarks used fixed filters or zoom-out; nobody has shown learned view selection working this well on HatefulIllusion. The main experiment is solid: message-level split, three seeds, ablations, error analysis, and a hyperparameter sweep all back the 93.2% closed-set claim. Original-view retrieval sits at 79.5%, filter baselines near chance, so the adaptive mechanism is doing real work. I also credit the authors for showing the gate's learned preferences by target type and visibility—that's the kind of analysis that makes the result believable.

The soft spots are proportionate. First and most important: the missing target-holdout number. The paper's own Limitations section flags it, and the stress-test note is right that without it we can't tell graceful degradation from collapse on unseen symbols. The authors should report balanced accuracy when entire symbol families are held out both from training and from the template bank. Second, the HC-Bench result in Table 2 is scored as correct if any of seven views answers, which is a best-of-seven oracle. That inflates the 98.2% and makes the comparison to zoom-out unequal. It's a secondary benchmark, so not fatal, but they should re-score with the adaptive decision. Third minor: hyperparameters and the view bank are tuned on the same benchmark; the one-at-a-time sweep is a start, but it doesn't fix the closed-set issue. The code release promise also needs a hash to be checkable.

Net: this is a useful, well-executed closed-set method with an unquantified generalization gap. It deserves a serious referee, but the referee's first request should be the unseen-template holdout. If that number is high, this is an important safety result. If it's near chance, the paper is still a reasonable demonstration of adaptive multi-view retrieval, just not a solution to the open-set problem it motivates.

Referee Report

3 major / 4 minor

Summary. The paper proposes Adaptive View Retrieval, a retrieve-and-calibrate framework for detecting hidden hateful messages in optical illusions. The method builds a seven-view perceptual bank (original, contrast, closure-edge, three figure-ground, low-pass), applies the same transforms to query images and to templates in a known-message library, embeds all views with a frozen CLIP encoder, uses an adaptive gate to weight views, retrieves the best-matching template, and calibrates a harmful/benign decision. On HatefulIllusion the authors report 93.2% balanced accuracy under a message-level held-out split, with ablations showing the view bank, adaptive gating, and joint retrieval/calibration training all contribute. The same view bank is transferred to IllusoryVQA (fine-tuned CLIP) and to HC-Bench (VLM multi-view questioning), where it outperforms single-view baselines. The paper includes component ablations, hyperparameter sensitivity, error analysis, and a public code release.

Significance. If the closed-set results are taken at face value, the paper demonstrates a substantial improvement over original-view classifiers and fixed-filter baselines on the HatefulIllusion benchmark, and the adaptive gate's dependence on target type and visibility is an interesting and interpretable finding. The component ablations (original view only 79.5%, uniform gate 89.1%, full model 93.2%) and the three-seed protocol support the central claim within the benchmark. The release of code and the explicit error analysis are also strengths. However, the practical significance is limited by the closed-set retrieval setting: the template bank contains a reference for every test message identity, and the paper acknowledges but does not quantify the drop under target holdout on unseen symbol families. The HC-Bench result is also scored as a best-of-seven oracle, which overstates the adaptive nature of the method. These concerns are addressable but are load-bearing for the generalization claims.

major comments (3)

[Discussion, Implications and Limitations; Eq. (1)] The method's decision rule is argmax over template bank M (Eq. 1), and the authors state that performance 'also drops under stricter target holdout on unseen symbol families' but never report the magnitude of that drop. Because the headline 93.2% is a closed-set retrieval result—every test message has a template in M—the practical benefit for real-world moderation, where new slangs and symbols emerge, is unquantified. I request a target-holdout experiment: train templates exclude an entire symbol family (or several), and report balanced accuracy on those unseen families. This is needed to distinguish graceful degradation from collapse and directly addresses the paper's stated limitation.
[Experiment 3, Table 2] The HC-Bench result is scored as 'counted correct if any transformed view yields a valid answer under the official rules.' This is a best-of-seven oracle, not an adaptive decision, and the comparison to original image or zoom-out (single views) is therefore not apples-to-apples. The reported 98.2% is an upper bound. Please also report per-view accuracies, and ideally an adaptive view-selection accuracy using the gate or a similar mechanism. Otherwise the claim that 'Adaptive View Retrieval' outperforms zoom-out on HC-Bench is overstated.
[Training and inference; Supplementary Table 2] The hyperparameter sweep in Supplementary Table 2 reports balanced accuracy on the test split. If the default hyperparameters were selected using test-split accuracy, the 93.2% is not a clean held-out estimate. Please clarify whether the sweep was conducted on the validation split or the test split. If the test split was used, rerun the sweep on the validation split (or report validation accuracy) and confirm the default remains the same.

minor comments (4)

[Method, Eq. (1)] The temperature τ in Eq. (1) is not defined. Is it a learned scalar initialized to some value? Please specify; this matters for the scale of the retrieval scores.
[Experiment 2, Table 1] The IllusoryVQA results are reported as single point estimates without variance across seeds. Since Experiment 1 uses three seeds, please add mean±std for Table 1 as well, particularly because the method uses full fine-tuning there.
[Supplementary Material] The main text references 'Supplementary Table 1' and 'Supplementary Figure 3,' but the supplementary file labels the relevant items as Table 3 and Figure 10. Please renumber consistently.
[Experiment 3] The HC-Bench experiment uses a fixed VLM and multi-view questioning, but does not use the gate or calibration head from Experiments 1–2. Please clarify in the text that this is a view-bank ablation under the SemVink protocol, not the full Adaptive View Retrieval architecture.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the 93.2% result is a supervised, closed-gallery retrieval evaluation and the paper explicitly disclaims open-vocabulary generalization.

full rationale

The derivation chain is self-contained. The retrieval score S_j(x)=τ∑_k α_k(x)⟨z_x^k,z_j^k⟩ and decision m_hat=argmax_{m_j∈M} S_j(x) are defined over a fixed template bank; the calibration head d(x) is a separately trained binary head. Each component is trained with cross-entropy and BCE losses on the train split and evaluated on a held-out message-level test split over three seeds, so no fitted quantity is renamed as a prediction. The template bank is a candidate gallery, and the paper explicitly acknowledges it is not open-vocabulary ('not open-vocabulary discovery of unseen hate strings'), so the closed-set nature is disclosed rather than hidden. There are no author self-citations in the reference list; the benchmarks cited (Qu et al. 2025; Rostamkhani et al. 2025; Li et al. 2025) are external. The one-at-a-time hyperparameter sweep reported on the test split is an evaluation-hygiene concern (tuning on test), not a circular reduction. Experiment 3's 'any transformed view yields a valid answer' scoring is a lenient oracle but is an explicit protocol choice, not an equivalence to the method's inputs. The paper's limitation statement that performance drops on unseen symbol families is a scope restriction; it lowers external validity but does not make the reported result equivalent to its inputs by construction. Verdict: no significant circularity.

Assumptions & free parameters 5 free parameters · 5 assumptions · 0 invented entities

The central claim is empirical and rests on the chosen view bank, the CLIP embedding space, and the template library as much as on the learned gate. The hand-picked transforms and learned temperature are the main fitted degrees of freedom; no new physical entities are introduced.

free parameters (5)

temperature τ = learned, not reported
Temperature scaling in the retrieval score S_j(x); fitted during training and central to retrieval scores.
detection loss weight λ_det = 0.5
BCE weight for the calibration head; chosen via a one-at-a-time sweep and near-insensitive per Table 4.
gate entropy weight λ_ent = 0.01
Regularizer to discourage view collapse; chosen by sweep.
learning rate / epochs = 1e-3 / 10
Selected by sweep; robust per Table 4.
view-bank transform constants = unspecified
Gaussian blur sizes, morphological closure/smoothing kernels, Otsu and low-pass parameters are hand-chosen and not given numerically; the method depends on them.

assumptions (5)

domain assumption Frozen CLIP ViT-B/32 embeddings of matched views are semantically aligned for retrieval.
Used for all retrieval scores; if CLIP cannot align transformed illusion views with template renderings, the method fails.
domain assumption The seven deterministic transforms reveal enough hidden structure for all target types and carriers.
The view bank is fixed by hand; if some hidden message is invisible under all seven views, retrieval cannot recover it (Figure 7 shows such a failure).
domain assumption The template bank M covers all test message identities.
Retrieval is argmax over M; the paper's stated limitation says open-vocabulary discovery is out of scope.
domain assumption HatefulIllusion labels and visibility annotations are correct.
All balanced-accuracy and slice numbers assume the benchmark labels are accurate.
standard math Otsu thresholding, Gaussian blur, and morphological operators behave as standard computer-vision primitives.
Used to construct figure-ground and edge views; no formal verification is provided.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Now You See the Hate: Adaptive View Retrieval for Hidden Hateful Illusions." pith.science (2026). https://pith.science/paper/NAKTUR42

@misc{pith2026260719061,
  author       = {Pith},
  title        = {Pith review of: Now You See the Hate: Adaptive View Retrieval for Hidden Hateful Illusions},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/NAKTUR42}},
  note         = {Machine review of arXiv:2607.19061}
}

read the original abstract

Hateful optical illusions expose a serious gap in current multimodal safety systems. On original-view hateful illusions, previous work shows that six moderation classifiers achieve at most 20.9 to 24.5% accuracy and nine state-of-the-art VLMs remain at or below 10.2% with illusion-aware prompting, leaving most hidden hate undetected. We formulate hidden hateful illusion detection as a perceptual retrieval problem and propose Adaptive View Retrieval. This retrieve-and-calibrate framework assembles a complementary view bank for the image and hidden-message templates, adaptively selects which views to trust, retrieves hidden-message identities, and calibrates whether the recovered evidence is harmful. On HatefulIllusion with a frozen CLIP encoder, Adaptive View Retrieval reaches 93.2% balanced accuracy on the held-out test split. It substantially outperforms original-view baselines and fixed single-transform filters across hate slangs, hate symbols, and visibility levels. The same design also surpasses official fine-tuned CLIP baselines, matches or exceeds human performance on IllusionMNIST, IllusionFashionMNIST, and IllusionAnimals, and outperforms zoom-out preprocessing on HC-Bench under the SemVink protocol. Together, these results show that robust multimodal moderation requires recovering hidden meaning before deciding whether it is harmful.

Figures

Figures reproduced from arXiv: 2607.19061 by the authors.

**Figure 2.** Overview of Adaptive View Retrieval. The template bank stores reference renderings of hidden messages, such as hate [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗

**Figure 3.** Example HatefulIllusion images. From left to right: [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗

Figures from the paper (8 more)

**Figure 4.** Figure 4: Mean gate composition for harmful targets by target type and visibility. Hate slangs rely predominantly on low-pass [PITH_FULL_IMAGE:figures/full_fig_p005_4.png]

**Figure 5.** Figure 5: Balanced accuracy on the held-out HatefulIllusion [PITH_FULL_IMAGE:figures/full_fig_p005_5.png]

**Figure 6.** Figure 6: Example hidden-hate images and their complementary view-bank outputs. The dominant view differs across examples, [PITH_FULL_IMAGE:figures/full_fig_p006_6.png]

**Figure 7.** Figure 7: A false negative on a low-visibility hate symbol. Contrast and low-pass reveal coarse layout, but figure-ground [PITH_FULL_IMAGE:figures/full_fig_p007_7.png]

**Figure 8.** Figure 8: Example IllusoryVQA images from IllusionM [PITH_FULL_IMAGE:figures/full_fig_p010_8.png]

**Figure 9.** Figure 9: Example HC-Bench images with hidden objects, [PITH_FULL_IMAGE:figures/full_fig_p010_9.png]

**Figure 10.** Figure 10: Moderation profile on the held-out HatefulIllusion test set. Each line reports one method on the full test set, the hard [PITH_FULL_IMAGE:figures/full_fig_p011_10.png]

**Figure 11.** Figure 11: Additional multi-view gallery examples on HatefulIllusion. Each row shows one hidden-hate image under the [PITH_FULL_IMAGE:figures/full_fig_p012_11.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

33 extracted references · 1 canonical work pages

[1]

ACM SIGGRAPH 2024 Conference Papers , pages=

Diffusion illusions: Hiding images in plain sight , author=. ACM SIGGRAPH 2024 Conference Papers , pages=

2024
[2]

Proceedings of the ACM on Web Conference 2025 , pages=

Illusioncaptcha: A captcha based on visual illusion , author=. Proceedings of the ACM on Web Conference 2025 , pages=

2025
[3]

Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , pages=

Visual anagrams: Generating multi-view optical illusions with diffusion models , author=. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , pages=
[4]

Proceedings of the IEEE/CVF conference on computer vision and pattern recognition , pages=

Convolutional neural networks can be deceived by visual illusions , author=. Proceedings of the IEEE/CVF conference on computer vision and pattern recognition , pages=
[5]

Vision Research , volume=

Color illusions also deceive CNNs for low-level vision tasks: Analysis and implications , author=. Vision Research , volume=. 2020 , publisher=

2020
[6]

Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , pages=

The art of deception: Color visual illusions and diffusion models , author=. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , pages=
[7]

Guan, Tianrui and Liu, Fuxiao and Wu, Xiyang and Xian, Ruiqi and Li, Zongxia and Liu, Xiaoyu and Wang, Xijun and Chen, Lichang and Huang, Furong and Yacoob, Yaser and others , booktitle=
[8]

Advances in Neural Information Processing Systems , volume=

Hidden in plain sight: Evaluating abstract shape recognition in vision-language models , author=. Advances in Neural Information Processing Systems , volume=

Show all 33 references

[9]

arXiv preprint arXiv:2603.24730 , year=

A Framework for Generating Semantically Ambiguous Images to Probe Human and Machine Perception , author=. arXiv preprint arXiv:2603.24730 , year=
[10]

Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing , pages=

SemVink: Advancing VLMs’ Semantic Understanding of Optical Illusions via Visual Global Thinking , author=. Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing , pages=

2025
[11]

Stevens' handbook of experimental psychology: Sensation and perception (vol

Perceptual organization in vision , author=. Stevens' handbook of experimental psychology: Sensation and perception (vol. 1, 3rd edn., pp. 177–234), John Wiley and Sons Inc, Hoboken, NJ, USA , year=
[12]

Proceedings of the Workshop on Cognitive Modeling and Computational Linguistics , pages=

Evaluating vision-language models on bistable images , author=. Proceedings of the Workshop on Cognitive Modeling and Computational Linguistics , pages=
[13]

Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , pages=

Do vision models perceive illusory motion in static images like humans? , author=. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , pages=
[14]

Proceedings of the Computer Vision and Pattern Recognition Conference , pages=

Illusory VQA: Benchmarking and enhancing multimodal models on visual Illusions , author=. Proceedings of the Computer Vision and Pattern Recognition Conference , pages=
[15]

Shahgir, Haz Sameen and Sayeed, Khondker Salman and Bhattacharjee, Abhik and Ahmad, Wasi Uddin and Dong, Yue and Shahriyar, Rifat , journal=
[16]

arXiv preprint arXiv:2506.05765 , year=

Do Large Vision-Language Models Distinguish between the Actual and Apparent Features of Illusions? , author=. arXiv preprint arXiv:2506.05765 , year=

arXiv
[17]

Sun, Xiaoxiao and Li, Mingyang and Yuan, Kun and Sun, Min Woo and Endo, Mark and Wu, Shengguang and Li, Changlin and Zhang, Yuhui and Wang, Zeyu and Yeung-Levy, Serena , booktitle=. Do
[18]

Perception , volume=

What are visual illusions? , author=. Perception , volume=. 2020 , publisher=

2020
[19]

arXiv preprint arXiv:2412.18613 , year=

The illusion-illusion: Vision language models see illusions where there are none , author=. arXiv preprint arXiv:2412.18613 , year=

arXiv
[20]

Proceedings of the IEEE/CVF International Conference on Computer Vision , pages=

Hate in Plain Sight: On the Risks of Moderating AI-Generated Hateful Illusions , author=. Proceedings of the IEEE/CVF International Conference on Computer Vision , pages=
[21]

Illusions in humans and

Yang, Jianyi and Ye, Junyi and Dash, Ankan and Wang, Guiling , journal=. Illusions in humans and
[22]

Proceedings of the 2023 conference on empirical methods in natural language processing , pages=

Grounding visual illusions in language: Do vision-language models perceive illusions like humans? , author=. Proceedings of the 2023 conference on empirical methods in natural language processing , pages=

2023
[23]

2025 , organization=

Zhang, Yiming and Zhang, Zicheng and Wei, Xinyi and Liu, Xiaohong and Zhai, Guangtao and Min, Xiongkuo , booktitle=. 2025 , organization=

2025
[24]

2026 , eprint=

A Survey of Multimodal Hallucination Evaluation and Detection , author=. 2026 , eprint=

2026
[25]

2026 , eprint=

Beyond Hate: Differentiating Uncivil and Intolerant Speech in Multimodal Content Moderation , author=. 2026 , eprint=

2026
[26]

SURE : Safety Understanding and Reasoning Enhancement for Multimodal Large Language Models

Gou, Yuxin and Dong, Xiaoning and Li, Qin and Gu, Shishen and Hong, Richang and Hu, Wenbo. SURE : Safety Understanding and Reasoning Enhancement for Multimodal Large Language Models. Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing. 2025. ...

2025 doi
[27]

2021 , journal=

Intriguing Properties of Vision Transformers , author=. 2021 , journal=

2021
[28]

Stochastic relativistic advection diffusion equation from the Metropolis algorithm , volume=

Başar, Gökçe and Bhambure, Jay and Singh, Rajeev and Teaney, Derek , year=. Stochastic relativistic advection diffusion equation from the Metropolis algorithm , volume=. Physical Review C , publisher=. doi:10.1103/physrevc.110.044903 , number=

doi
[29]

2021 , eprint=

Better Aggregation in Test-Time Augmentation , author=. 2021 , eprint=

2021
[30]

Proceedings of the IEEE/CVF international conference on computer vision , pages=

Better aggregation in test-time augmentation , author=. Proceedings of the IEEE/CVF international conference on computer vision , pages=
[31]

2022 , eprint=

Trion gas, electron-hole liquid, and metal-insulator transition in doped heterostructures based on transition metal dichalcogenides , author=. 2022 , eprint=

2022
[32]

Perceptual grouping and figure--ground organization

A century of Gestalt psychology in visual perception: I. Perceptual grouping and figure--ground organization. , author=. Psychological bulletin , volume=. 2012 , publisher=

2012
[33]

, author=

Laws of organization in perceptual forms. , author=. 1938 , journal=

1938

Pith tools

Reviewed August 1, 2026 · model on record in the stance chip above.

[1] [1]

ACM SIGGRAPH 2024 Conference Papers , pages=

Diffusion illusions: Hiding images in plain sight , author=. ACM SIGGRAPH 2024 Conference Papers , pages=

2024

[2] [2]

Proceedings of the ACM on Web Conference 2025 , pages=

Illusioncaptcha: A captcha based on visual illusion , author=. Proceedings of the ACM on Web Conference 2025 , pages=

2025

[3] [3]

Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , pages=

Visual anagrams: Generating multi-view optical illusions with diffusion models , author=. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , pages=

[4] [4]

Proceedings of the IEEE/CVF conference on computer vision and pattern recognition , pages=

Convolutional neural networks can be deceived by visual illusions , author=. Proceedings of the IEEE/CVF conference on computer vision and pattern recognition , pages=

[5] [5]

Vision Research , volume=

Color illusions also deceive CNNs for low-level vision tasks: Analysis and implications , author=. Vision Research , volume=. 2020 , publisher=

2020

[6] [6]

Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , pages=

The art of deception: Color visual illusions and diffusion models , author=. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , pages=

[7] [7]

Guan, Tianrui and Liu, Fuxiao and Wu, Xiyang and Xian, Ruiqi and Li, Zongxia and Liu, Xiaoyu and Wang, Xijun and Chen, Lichang and Huang, Furong and Yacoob, Yaser and others , booktitle=

[8] [8]

Advances in Neural Information Processing Systems , volume=

Hidden in plain sight: Evaluating abstract shape recognition in vision-language models , author=. Advances in Neural Information Processing Systems , volume=