Pith. sign in

REVIEW 5 major objections 6 minor 29 references

Targeted False Positive Synthesis via Detector-guided Adversarial Diffusion Attacker for Robust Polyp Detection

T0 review · 5 major / 6 minor · reviewed 2026-08-06 · deepseek-v4-flash

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

Pith's one-line read Synthesizing detector-confusing negatives—rather than more polyp positives—teaches YOLO and DETR to reject polyp lookalikes, improving F1-score by at least 2.6% on Kvasir and 2.7% on an in-house dataset.

desk verdict Novel combination of background-only diffusion and detector-guided attacks for synthesizing false positives; the idea is promising but the evaluation is not yet auditable because the attack region b is undisclosed and alpha is tuned on the test split. read the letter →

arxiv 2506.18134 v1 pith:KCEHVN7Y submitted 2025-06-22 cs.CV

Quan Zhou , Gan Luo , Qiang Hu , Qingyong Zhang , Jinhua Zhang , Yinjiao Tian , Qiang Li , Zhiwei Wang This is my paper · ORCID
classification cs.CV
keywords adversarialdiffusionfalsepositivesynthesispolypdetectiondataaugmentationdenoisingprobabilisticmodeldetectorguidancecolonoscopycolorectalcancerscreening
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

The paper argues that the false positives that plague automated polyp detection can be treated as a synthesis problem rather than a filtering problem. It builds a background-only diffusion model that learns what normal colonoscopy tissue looks like, then adds a detector-guided adversarial module that nudges the denoising process toward polyp-like patterns that fool a trained detector. The resulting images are used as extra training negatives, teaching YOLO and DETR detectors to reject polyp-lookalikes instead of flagging them. On the Kvasir benchmark and an in-house clinical dataset, the augmented detectors improve F1-score by at least 2.6 and 2.7 percentage points over baselines, which matters because reducing false positives lowers unnecessary interventions and clinician alarm fatigue.

What carries the argument

The central object is the Detector-guided Adversarial Diffusion Attacker (DADA), which maintains a perturbation variable $\eta_t$ injected at each denoising step of a background-only DDPM. BG-De masks out polyp bounding boxes during training so it models only non-polyp backgrounds, and DADA backpropagates the detector's detection loss through the denoising trajectory, updating $\eta_t$ by sign-gradient descent to steer generation toward patterns that trigger a false positive in a user-defined box $b$. An inpainting step keeps the surrounding context from the real image so the synthesized confusers stay anatomically coherent. Together the modules convert ordinary backgrounds into detector-confusing false positives.

What would settle it

Run a DADA-augmented detector on a held-out set of real, polyp-free colonoscopy frames recorded with a different endoscope; the central claim fails if the detector's false-positive rate is not lower than that of a baseline detector trained without DADA data, because the synthetic negatives would have taught it only the teacher's specific confusers, not the real ones.

Watch

Extended reading notes

Core claim

The central claim is that synthesizing high-value false positives—negatives that resemble polyps closely enough to provoke a detector response—is a more effective use of generative data augmentation than generating more polyp positives or generic backgrounds. The paper demonstrates this with DADA, which couples a negative-centric DDPM called BG-De with an adversarial attacker that maximizes the detector's classification and localization loss in a user-specified region. The mechanism is the first application of adversarial diffusion to lesion detection, and the experimental payoff is that both a CNN detector (YOLO) and a transformer detector (DETR) gain precision and recall when trained on the synthesized confusers, with the largest gains on detector precision.

Load-bearing premise

The load-bearing premise is that the adversarial patterns found against the teacher detector in a user-chosen attack region also reflect the real false positives that occur in the held-out test distribution; if the attack region or the teacher's failure modes are unrepresentative, the precision gains will not generalize.

Editorial extensions

If this is right

  • Detectors trained with DADA-augmented data improve F1 by at least 2.6% on Kvasir and 2.7% on the in-house dataset over baselines, for both YOLO and DETR backbones.
  • Both components are necessary: ablations show that BG-De alone gives limited gains, DADA alone degrades performance by introducing label noise, and only their joint use yields the best results.
  • Larger perturbation step sizes $\alpha$ raise the false-positive generation rate but lower image fidelity, and an intermediate $\alpha=0.003$ gives the best detector F1-score.
  • The framework requires no extra data collection because it uses only polyp images with standard ground-truth bounding boxes, which are already available in detection datasets.

Reading between the lines

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

  • Beyond the paper: if the adversarial patterns transfer across endoscope types and bowel-preparation protocols, DADA-style synthesis could be used as a controllable augmentation tool for other lesion-detection tasks where false positives are costly, such as lung nodules or skin lesions.
  • Beyond the paper: the user-specified attack region $b$ could be chosen adversarially at inference time to probe a deployed detector for hallucination-prone areas, turning the attacker into a real-time reliability monitor.
  • Beyond the paper: the paper's FPGR and FID measurements suggest a tunable difficulty-realism trade-off; future work could build a calibration curve that lets clinicians set the perturbation strength to a clinically acceptable false-positive rate.
  • Beyond the paper: the teacher-detector dependence implies that the method may need re-tuning when the final deployment detector differs from the attacked detector, since the synthesized confusers are tailored to the teacher's decision boundary.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

5 major / 6 minor

Summary. The paper proposes a detector-guided adversarial diffusion framework (DADA) for synthesizing targeted false-positive samples to improve polyp detection. The method couples a background-only denoising diffusion model (BG-De), trained with a regional noise-matching loss that masks polyp ground-truth boxes, with an adversarial attacker that perturbs the denoising trajectory at each step to induce a pre-trained detector to predict a false positive in a user-specified bounding box. An inpainting strategy preserves real-image context outside the attacked region. Experiments on Kvasir and an in-house dataset with YOLO and DETR detectors report F1 improvements of at least 2.6 and 2.7 points over baselines, plus ablations showing that both BG-De and DADA are needed. The paper also studies the perturbation step size alpha via FID and a false-positive generation rate metric.

Significance. If validated, the work addresses a practically important problem: reducing false positives in computer-aided polyp detection, where excessive alerts disrupt clinical workflow. The formulation in Eqs. (1)-(6) is coherent, and the two-component design is well motivated and clearly supported by the ablation of Table 2 as an internal consistency check. The paper also includes experiments on two datasets and two detector architectures, compares against two adversarial-attack and two inpainting baselines, and announces public code. These are genuine strengths. However, the central empirical claim is compromised by the absence of a protocol for selecting the attack region b and by test-set-informed hyperparameter selection, as detailed below. The reported gains are therefore not yet fully auditable.

major comments (5)
  1. [Sec. 2.2-2.3 and Sec. 3.2] The central claim depends on the user-defined bounding box b, but the paper never specifies how b is selected on either dataset. b is the illusory ground-truth box in Eq. (4) and the inpainted region in Eq. (6), so it fully determines the spatial location and visual content of each generated false positive. There is no statement that b is chosen from training or validation information only, and no description of the selection protocol (e.g., random background crops, a fixed grid, or boxes placed at detector failure locations on the training set). If b was chosen by inspecting where the baseline detector produces false positives on the test split, the reported F1 gains in Table 1 would reflect test-information leakage rather than a property of the synthesis method. The authors must specify the exact protocol and confirm that no test-set information is used.
  2. [Sec. 3.5 and Table 3] The perturbation step size alpha is selected by maximizing F1 on the test set. Table 3 reports F1 values for each alpha, and the chosen alpha=0.003 yields exactly the 'Ours' F1 values in Table 1 (0.969 on Kvasir, 0.891 on the in-house dataset). This is test-set hyperparameter tuning, which inflates reported performance and makes the comparison with baselines unfair, since the baselines are not tuned in the same way. The authors should select alpha on the validation split (the 8:1:1 split already provides one) and report the resulting test-set numbers. Additionally, no error bars, confidence intervals, or multiple-seed runs are reported anywhere in Tables 1-3, so it is impossible to assess whether the 2.6/2.7-point gaps exceed run-to-run variance.
  3. [Table 1 vs. Table 2] The 'no DADA, no BG-De' row in the ablation Table 2 should reproduce the baseline in Table 1 under the same setting (no augmented data), but it does not: YOLO F1 is 0.937 in Table 2 versus 0.941 in Table 1, and DETR F1 is 0.897 versus 0.901. This discrepancy is unexplained and undermines the internal consistency of the ablation. The authors should clarify what differs between the two configurations (e.g., retraining with a different seed or a different data split) and report consistent baseline numbers.
  4. [Sec. 3.3] No baseline that augments the training set with real negative images or ordinary synthesized negatives at matched volume is included. The comparisons are limited to adversarial-attack methods (APGD, FAB) and inpainting methods (Repaint, LaMa). Without a simple real-negative augmentation control, the F1 gains in Table 1 may be explained by adding more negative training examples in general rather than by the targeted false-positive synthesis mechanism specifically. Adding such a baseline (e.g., augmenting each training image with a real non-polyp background patch or a randomly sampled negative image) would directly test the claimed mechanism.
  5. [Sec. 3.5] The False Positive Generation Rate (FPGR) metric used to justify the choice of alpha is not defined with enough precision. 'The proportion of generated images causing false positives' requires specifying which detector is used, what confidence threshold is applied, and whether the evaluation is done on the teacher detector or the final trained detector. Without this information, the metric is not reproducible and cannot support the hyperparameter comparison in Table 3.
minor comments (6)
  1. [Table 1] The table caption contains a typo: 'T able 1' should be 'Table 1'.
  2. [Table 1] The method label 'F AB' should be 'FAB' for consistency with the reference list.
  3. [Fig. 1] The detector is labeled 'Detr' in the figure; for consistency with the text it should be 'DETR'.
  4. [Sec. 2.1] The sentence 'with1assigned to pixels inside' is missing a whitespace after 'with'; it should read 'with 1 assigned'.
  5. [Sec. 3.1 and Table 1] The text consistently calls the second dataset 'in-house dataset', but the caption of Table 1 calls it 'private dataset'; the terminology should be unified.
  6. [Sec. 3.2] The paper states 'we set the denoising steps to 1,000' for inference but does not specify the number of sampling steps used for the inpainting variants; clarifying this would improve reproducibility.

Circularity Check

0 steps flagged · score 2.0 of 10

No significant circularity: the synthesis pipeline is standard adversarial augmentation; only a minor non-load-bearing self-citation prevents a perfect 0.

full rationale

The central claim is that DADA-generated false positives improve detector F1 by 2.6% and 2.7% on Kvasir and in-house data, respectively. This is an empirical result obtained by training detectors on original plus augmented data and evaluating on a held-out test split; it is not derived from the equations defining BG-De or DADA. The generation pipeline (Eqs. 1, 3, 5, 6) uses a pretrained detector to guide synthesis of adversarial negatives, which is a standard adversarial-training/hard-negative-mining procedure rather than a definitional equivalence between input and output. The attack region b is a user-specified input, not a fitted parameter, and the paper's reported F1 values are measured outcomes, not quantities forced by construction. The only self-citation is Ref. [10] by overlapping authors (Hu, Li, Wang), cited in the introduction for reflection artifacts; it is contextual and not load-bearing. The hyperparameter search over alpha in Sec. 3.5, if performed on validation, is standard; the paper does not explicitly state the split, which is a reproducibility/soundness concern rather than circularity. Similarly, the missing protocol for selecting b could indicate test-informed region choice and would be data leakage, but no circular reduction is exhibited. Overall, the derivation chain is self-contained against external benchmarks, so the circularity score is low.

Assumptions & free parameters 2 free parameters · 3 assumptions · 0 invented entities

No new physical or conceptual entities are introduced; all components are existing models (DDPM, DETR/YOLO) and standard losses. The method introduces a module (DADA) but it is an algorithm, not an invented entity.

free parameters (2)
  • Perturbation step size alpha = 0.003
    Selected in Sec 3.5 by maximizing detector F1 on the test sets (Kvasir F1 0.969, in-house F1 0.891). This is test-set hyperparameter fitting.
  • Illusory bounding box b (attack region) = Not specified; user-defined
    The location and size of the region used to generate false positives is described as 'user-defined' (Sec 2.2-2.3) but the experimental protocol for choosing b is not given, which could affect the difficulty and diversity of generated negatives.
assumptions (3)
  • standard math DDPM forward and reverse processes as in Ho et al. [19] are correct and applicable.
    Used in Sec 2.1 to define BG-De training and sampling.
  • domain assumption Gradients from the pre-trained detector's loss on an 'illusory' box provide a signal that yields transferable false positives.
    Sec 2.2 optimizes eta to minimize Ldet; the paper assumes this produces useful negatives rather than merely adversarial artifacts.
  • domain assumption The user-specified region b can be chosen without biasing evaluation.
    Sec 2.3 uses mb to inpaint; if b were chosen based on test-set false positives, results would be inflated. The paper does not state the selection rule.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Targeted False Positive Synthesis via Detector-guided Adversarial Diffusion Attacker for Robust Polyp Detection." pith.science (2026). https://pith.science/paper/KCEHVN7Y

@misc{pith2026250618134,
  author       = {Pith},
  title        = {Pith review of: Targeted False Positive Synthesis via Detector-guided Adversarial Diffusion Attacker for Robust Polyp Detection},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/KCEHVN7Y}},
  note         = {Machine review of arXiv:2506.18134}
}
read the original abstract

Polyp detection is crucial for colorectal cancer screening, yet existing models are limited by the scale and diversity of available data. While generative models show promise for data augmentation, current methods mainly focus on enhancing polyp diversity, often overlooking the critical issue of false positives. In this paper, we address this gap by proposing an adversarial diffusion framework to synthesize high-value false positives. The extensive variability of negative backgrounds presents a significant challenge in false positive synthesis. To overcome this, we introduce two key innovations: First, we design a regional noise matching strategy to construct a negative synthesis space using polyp detection datasets. This strategy trains a negative-centric diffusion model by masking polyp regions, ensuring the model focuses exclusively on learning diverse background patterns. Second, we introduce the Detector-guided Adversarial Diffusion Attacker (DADA) module, which perturbs the negative synthesis process to disrupt a pre-trained detector's decision, guiding the negative-centric diffusion model to generate high-value, detector-confusing false positives instead of low-value, ordinary backgrounds. Our approach is the first to apply adversarial diffusion to lesion detection, establishing a new paradigm for targeted false positive synthesis and paving the way for more reliable clinical applications in colorectal cancer screening. Extensive results on public and in-house datasets verify the superiority of our method over the current state-of-the-arts, with our synthesized data improving the detectors by at least 2.6% and 2.7% in F1-score, respectively, over the baselines. Codes are at https://github.com/Huster-Hq/DADA.

Figures

Figures reproduced from arXiv: 2506.18134 by the authors.

Figure 1
Figure 1. The inference pipeline consists of three key modules: the background-only de￾noiser (BG-De), a well-trained polyp detector (Detr), and the detector-guided adver￾sarial diffusion attacker (DADA). Inpainting is used to generate new negative samples in a user-specified region of the real image. BG-De samples from the pure negative dis￾tribution, while DADA guides the sampling toward visual patterns that elicit a positi… view at source ↗
Figure 2
Figure 2. Visualizations of synthesis images by different methods. The white bounding boxes indicate pre-defined attack/inpainting regions. For clarity, we visualize difference maps D between synthesis images and original images [PITH_FULL_IMAGE:figures/full_fig_p007_2.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

29 extracted references · 24 canonical work pages

  1. [1]

    Colorectal cancer: a review.Int J Res Med Sci, 5(11):4667, 2017

    Juan José Granados-Romero, Alan Isaac Valderrama-Treviño, Ericka Hazzel Contreras-Flores, Baltazar Barrera-Mera, Miguel Herrera Enríquez, Karen Uriarte- Ruíz, Jesús Carlos Ceballos-Villalba, Aranza Guadalupe Estrada-Mata, Cristopher Alvarado Rodríguez, and Gerardo Arauz-Peña. Colorectal cancer: a review.Int J Res Med Sci, 5(11):4667, 2017

    work page 2017

  2. [2]

    Real-time polyp detection in colonoscopy using lightweight transformer

    Youngbeom Yoo, Jae Young Lee, Dong-Jae Lee, Jiwoon Jeon, and Junmo Kim. Real-time polyp detection in colonoscopy using lightweight transformer. In2024 IEEE/CVF Winter Conference on Applications of Computer Vision (WACV), pages 7794–7804. IEEE, 2024

    work page 2024

  3. [3]

    A robust real-time deep learning based auto- matic polyp detection system.Computers in Biology and Medicine, 134:104519, 2021

    Ishak Pacal and Dervis Karaboga. A robust real-time deep learning based auto- matic polyp detection system.Computers in Biology and Medicine, 134:104519, 2021

    work page 2021

  4. [4]

    Colon polyp detection and segmentation based on improved mrcnn.IEEE Transactions on Instrumentation and Measurement, 70:1–10, 2020

    Xiaoyong Yang, Qianxing Wei, Changhe Zhang, Kaibo Zhou, Li Kong, and Weiwei Jiang. Colon polyp detection and segmentation based on improved mrcnn.IEEE Transactions on Instrumentation and Measurement, 70:1–10, 2020

    work page 2020

  5. [5]

    A self- attention based faster r-cnn for polyp detection from colonoscopy images.Biomed- ical Signal Processing and Control, 70:103019, 2021

    Bo-Lun Chen, Jing-Jing Wan, Tai-Yue Chen, Yong-Tao Yu, and Min Ji. A self- attention based faster r-cnn for polyp detection from colonoscopy images.Biomed- ical Signal Processing and Control, 70:103019, 2021

    work page 2021

  6. [6]

    Polyp detection on video colonoscopy using a hybrid 2d/3d cnn.Medical Image Analysis, 82:102625, 2022

    JuanaGonzález-BuenoPuyal,PatrickBrandao,OmerFAhmad,KanwalKBhatia, DanielToth,RawenKader,LaurenceLovat,PeterMountney,andDanailStoyanov. Polyp detection on video colonoscopy using a hybrid 2d/3d cnn.Medical Image Analysis, 82:102625, 2022

    work page 2022

  7. [7]

    Polyp tracking in video colonoscopy using optical flow with an on-the-fly trained cnn

    He Zheng, Hanbo Chen, Junzhou Huang, Xuzhi Li, Xiao Han, and Jianhua Yao. Polyp tracking in video colonoscopy using optical flow with an on-the-fly trained cnn. In2019 IEEE 16th international symposium on biomedical imaging (ISBI 2019), pages 79–82. IEEE, 2019

    work page 2019

  8. [8]

    An end-to-end tracking method for polyp detectors in colonoscopy videos.Artificial intelligence in medicine, 131:102363, 2022

    TaoYu,NeLin,XuZhang,YanqiPan,HuiyiHu,WenfangZheng,JiquanLiu,Weil- ing Hu, Huilong Duan, and Jianmin Si. An end-to-end tracking method for polyp detectors in colonoscopy videos.Artificial intelligence in medicine, 131:102363, 2022

    work page 2022

Show all 29 references
  1. [9]

    Understanding the effects of artifacts on automated polyp detection and incorporating that knowledge via learning without forgetting.arXiv preprint arXiv:2002.02883, 2020

    Maxime Kayser, Roger D Soberanis-Mukul, Anna-Maria Zvereva, Peter Klare, Nassir Navab, and Shadi Albarqouni. Understanding the effects of artifacts on automated polyp detection and incorporating that knowledge via learning without forgetting.arXiv preprint arXiv:2002.02883, 20...

    2002 arXiv

  2. [10]

    Sali:Short-termalignmentandlong-terminteractionnetworkforcolonoscopyvideo polyp segmentation

    Qiang Hu, Zhenyu Yi, Ying Zhou, Fang Peng, Mei Liu, Qiang Li, and Zhiwei Wang. Sali:Short-termalignmentandlong-terminteractionnetworkforcolonoscopyvideo polyp segmentation. InInternational Conference on Medical Image Computing and Computer-Assisted Intervention, pages 531–541....

    2024

  3. [11]

    Multi-scale hybrid network for polyp detection in wireless capsule endoscopy and colonoscopy images.Diagnostics, 12(8):2030, 2022

    Meryem Souaidi and Mohamed El Ansari. Multi-scale hybrid network for polyp detection in wireless capsule endoscopy and colonoscopy images.Diagnostics, 12(8):2030, 2022

    2022

  4. [12]

    Examining the effect of synthetic data augmentation in polyp detection and segmentation.International Journal of Computer Assisted Radiology and Surgery, 17(7):1289–1302, 2022

    Prince Ebenezer Adjei, Zenebe Markos Lonseko, Wenju Du, Han Zhang, and Nini Rao. Examining the effect of synthetic data augmentation in polyp detection and segmentation.International Journal of Computer Assisted Radiology and Surgery, 17(7):1289–1302, 2022

    2022

  5. [13]

    A full data augmentation pipeline for small object detec- tion based on generative adversarial networks.Pattern Recognition, 133:108998, 2023

    Brais Bosquet, Daniel Cores, Lorenzo Seidenari, Víctor M Brea, Manuel Mucientes, and Alberto Del Bimbo. A full data augmentation pipeline for small object detec- tion based on generative adversarial networks.Pattern Recognition, 133:108998, 2023

    2023

  6. [14]

    Simple u-net based synthetic polyp image generation: Polyp to negative and negative to polyp

    Hemin Ali Qadir, Ilangko Balasingham, and Younghak Shin. Simple u-net based synthetic polyp image generation: Polyp to negative and negative to polyp. Biomedical Signal Processing and Control, 74:103491, 2022

    2022

  7. [15]

    Controlpolypnet: Towards controlled colon polyp synthesis for improved polyp segmentation

    Vanshali Sharma, Abhishek Kumar, Debesh Jha, MK Bhuyan, Pradip K Das, and Ulas Bagci. Controlpolypnet: Towards controlled colon polyp synthesis for improved polyp segmentation. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 2325–2334, 2024

    2024

  8. [16]

    Comparing the number and relevance of false activations between 2 artificial intelligence computer-aided detection systems: the noise study

    Marco Spadaccini, Cesare Hassan, Ludovico Alfarone, Leonardo Da Rio, Roberta Maselli,SilviaCarrara,PieraAlessiaGaltieri,GaiaPellegatta,AlessandroFugazza, Glenn Koleth, et al. Comparing the number and relevance of false activations between 2 artificial intelligence computer-aid...

    2022

  9. [17]

    Benchmarking defini- tions of false-positive alerts during computer-aided polyp detection in colonoscopy

    Erik A Holzwanger, Mohammad Bilal, Jeremy R Glissen Brown, Shailendra Singh, Aymeric Becq, Kenneth Ernest-Suarez, and Tyler M Berzin. Benchmarking defini- tions of false-positive alerts during computer-aided polyp detection in colonoscopy. Endoscopy, 53(09):937–940, 2021

    2021

  10. [18]

    Addressing false-positive findings with arti- ficial intelligence for polyp detection.Endoscopy, 53(09):941–942, 2021

    Yuichi Mori and Michael Bretthauer. Addressing false-positive findings with arti- ficial intelligence for polyp detection.Endoscopy, 53(09):941–942, 2021

    2021

  11. [19]

    Denoising diffusion probabilistic mod- els.Advances in neural information processing systems, 33:6840–6851, 2020

    Jonathan Ho, Ajay Jain, and Pieter Abbeel. Denoising diffusion probabilistic mod- els.Advances in neural information processing systems, 33:6840–6851, 2020

    2020

  12. [20]

    Towardsdeeplearningmodelsresistanttoadversarialattacks.arXiv preprint arXiv:1706.06083, 2017

    Aleksander Madry, Aleksandar Makelov, Ludwig Schmidt, Dimitris Tsipras, and AdrianVladu. Towardsdeeplearningmodelsresistanttoadversarialattacks.arXiv preprint arXiv:1706.06083, 2017

    2017 arXiv

  13. [21]

    Black-box adver- sarial attacks with limited queries and information

    Andrew Ilyas, Logan Engstrom, Anish Athalye, and Jessy Lin. Black-box adver- sarial attacks with limited queries and information. InInternational conference on machine learning, pages 2137–2146. PMLR, 2018

    2018

  14. [22]

    The hungarian method for the assignment problem.Naval research logistics quarterly, 2(1-2):83–97, 1955

    Harold W Kuhn. The hungarian method for the assignment problem.Naval research logistics quarterly, 2(1-2):83–97, 1955

    1955

  15. [23]

    End-to-end object detection with transformers

    Nicolas Carion, Francisco Massa, Gabriel Synnaeve, Nicolas Usunier, Alexander Kirillov, and Sergey Zagoruyko. End-to-end object detection with transformers. InEuropean conference on computer vision, pages 213–229. Springer, 2020

    2020

  16. [24]

    Kvasir-seg: A segmented polyp dataset

    DebeshJha,PiaHSmedsrud,MichaelARiegler,PålHalvorsen,ThomasDeLange, Dag Johansen, and Håvard D Johansen. Kvasir-seg: A segmented polyp dataset. In MultiMedia modeling: 26th international conference, MMM 2020, Daejeon, South Korea, January 5–8, 2020, proceedings, part II 26, pag...

    2020

  17. [25]

    ultra- lytics/yolov5: v6

    Glenn Jocher, Alex Stoken, Ayush Chaurasia, Jirka Borovec, Yonghye Kwon, Kalen Michael, Liu Changyu, Jiacong Fang, Piotr Skalski, Adam Hogan, et al. ultra- lytics/yolov5: v6. 0-yolov5n’nano’models, roboflow integration, tensorflow export, opencv dnn support.Zenodo, 2021

    2021

  18. [26]

    Reliable evaluation of adversarial robustness with an ensemble of diverse parameter-free attacks

    Francesco Croce and Matthias Hein. Reliable evaluation of adversarial robustness with an ensemble of diverse parameter-free attacks. InInternational conference on machine learning, pages 2206–2216. PMLR, 2020

    2020

  19. [27]

    Minimally distorted adversarial examples with a fast adaptive boundary attack

    Francesco Croce and Matthias Hein. Minimally distorted adversarial examples with a fast adaptive boundary attack. InInternational conference on machine learning, pages 2196–2205. PMLR, 2020

    2020

  20. [28]

    Repaint: Inpainting using denoising diffusion probabilistic mod- els

    Andreas Lugmayr, Martin Danelljan, Andres Romero, Fisher Yu, Radu Timofte, and Luc Van Gool. Repaint: Inpainting using denoising diffusion probabilistic mod- els. InProceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 11461–11471, 2022

    2022

  21. [29]

    Resolution-robust large mask inpainting with fourier con- volutions

    Roman Suvorov, Elizaveta Logacheva, Anton Mashikhin, Anastasia Remizova, Ar- senii Ashukha, Aleksei Silvestrov, Naejin Kong, Harshith Goka, Kiwoong Park, and Victor Lempitsky. Resolution-robust large mask inpainting with fourier con- volutions. InProceedings of the IEEE/CVF wi...

    2022

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.