Pith. sign in

REVIEW 3 major objections 5 minor 63 references

Augmentation Matters: A Mix-Paste Method for X-Ray Prohibited Item Detection under Noisy Annotations

T0 review · 3 major / 5 minor · reviewed 2026-08-10 · deepseek-v4-flash

Pith's one-line read The paper shows that blending same-label item patches into a mixed patch and pasting it back makes X-ray prohibited-item detectors robust to both category-label and bounding-box noise.

desk verdict A plausible and well-tested augmentation method for noisy X-ray detection, but the headline numbers are all under synthetic random noise; the structured-noise gap is real and should be addressed before publication. read the letter →

arxiv 2501.01733 v1 pith:VCCRAIBG submitted 2025-01-03 cs.CV cs.AI

classification cs.CVcs.AI
keywords X-rayprohibiteditemdetectionnoisyannotationsdataaugmentationMix-Pastelarge-losssuppressionobjectcategorylabelnoiseboundingbox
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 argues that data augmentation can absorb a large share of annotation noise in X-ray prohibited-item detection, even when both category labels and bounding boxes are corrupted. The proposed Mix-Paste method mixes K item patches that carry the same category label and pastes the blended patch back into the image, which raises the chance that the boxed region really contains an item of that class. A companion item-based large-loss suppression (LLS) rule drops the classification losses of plausible detections whose predicted category contradicts the noisy ground-truth label. On OPIXray with 60% category noise and 60% box noise, the combination lifts mAP@.5 from 56.7% to 81.8% and mAP@[.5,.95] from 18.4% to 33.7%, and similar gains appear on PIDray and, more modestly, on noisy MS-COCO. The paper's central claim is that this inference-free, plug-and-play recipe is a general way to train robust detectors without needing clean data or noise-rate estimates.

What carries the argument

The central object is the mixed patch $\hat{B} = \alpha \odot B_a + \sum_{n=2}^K \frac{1-\alpha}{K-1} \odot \mathrm{resize}(B_n)$, where $B_a$ is the original patch, the $B_n$ are randomly chosen patches with the same category label, and $\alpha$ is an edge-smoothing mask. This patch carries the probability amplification $1-P_c^K$ and is pasted back at the original box location. The second mechanism is item-based LLS, which splits predictions into negative, background, positive, and potentially-positive categories and discards the classification loss of the potentially-positive ones (high IoU, non-background label differing from ground truth). Together the machinery converts noisy labels into a training signal dominated by correct items while suppressing the loss spikes caused by extra blended items.

What would settle it

Re-annotate a subset of OPIXray or PIDray with clean labels, inject structured noise such as confusion concentrated between visually similar categories and boxes consistently inflated in one direction, and check whether Mix-Paste plus LLS keeps its advantage; if the gain collapses, the independence assumption behind $1-P_c^K$ is the weak link.

Watch

Extended reading notes

Core claim

Under noisy annotations, a single ground-truth box with label $c$ contains a correctly labeled item with probability roughly $1-P_c$ when the category noise rate is $P_c$. If K patches all labeled $c$ are mixed, the probability that the resulting region does not contain a correct item of class $c$ falls to $P_c^K$, so the chance the region is genuinely correct rises to $1-P_c^K$; the same reasoning applies to bounding-box noise with rate $P_b$, giving $1-P_b^K$. Mix-Paste implements this by resizing K same-label patches to one size, linearly blending them with an edge-smoothing mask, and pasting the result over the original box. Because the blended patch can also introduce extra items whose labels are not recorded, LLS identifies high-IoU predictions whose label disagrees with the ground-truth label and excludes their classification losses from the training objective. The paper reports that this two-part recipe consistently outperforms prior robust-detection and label-noise methods on OPIXray and PIDray across noise rates from 20% to 60%, and transfers to MS-COCO.

Load-bearing premise

The synthetic noise used in experiments, with random label replacement and uniformly perturbed boxes, is assumed to reproduce how real X-ray annotations get corrupted, and if real mistakes are systematic rather than random, the measured gains may not transfer.

Editorial extensions

If this is right

  • At 60% category noise and 60% box noise on OPIXray, Mix-Paste with LLS raises mAP@.5 from 56.7% to 81.8% and mAP@[.5,.95] from 18.4% to 33.7%, a gain of 25.1 and 15.3 points over the Faster R-CNN baseline.
  • Similar gains hold on PIDray, where the method improves average mAP@[.5,.95] by 17.0 points at the 60% noise level across easy, hard, and hidden test splits.
  • The recipe is plug-and-play: applying Mix-Paste alone improves two-stage, one-stage, and X-ray-specific detectors, and adding LLS gives a further gain over Mix-Paste alone.
  • On noisy MS-COCO, the same augmentation improves over the baseline by about 2.1 mAP@.5 and 2.9 mAP@[.5,.95] at 60% noise, showing the idea is not restricted to X-ray imagery.
  • The paper also reports that mixing with large K (4 or 5) or applying Mix-Paste to essentially all training images hurts performance; K=2 or 3 and a 0.6 mixing probability are the stable choices.

Reading between the lines

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

  • Beyond the paper's random-noise protocol, the probability gain $1-P_c^K$ presumes that label mistakes are independent; under systematic confusion of visually similar knife types, mixing may consolidate the wrong class rather than dilute it.
  • A natural next experiment would separate Mix-Paste's noise-suppression effect from its overlap-mimicking effect by running it on a fully clean, non-overlapping dataset; the original-OPIXray result suggests both effects are present.
  • Because Mix-Paste is a data-level operation, it should combine with label-refinement and loss-correction methods; the paper does not test such a combination.
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.

Referee Report

3 major / 5 minor

Summary. The paper proposes Mix-Paste, a data augmentation method for training X-ray prohibited-item detectors under noisy annotations (category label noise and bounding-box noise). For each ground-truth item patch, Mix-Paste randomly selects K same-label patches from other images, blends them via an edge-smoothing mask, and pastes the blend back into the original bounding box. The authors argue that mixing increases the probability that the augmented patch contains a correctly labeled item (1 - P_c^K) and, analogously, that it improves box accuracy (1 - P_b^K). They also introduce an item-based large-loss suppression (LLS) rule that discards classification losses for high-IoU predictions whose predicted label differs from the ground-truth label. Experiments on OPIXray, PIDray, and noisy MS-COCO show that Mix-Paste plus LLS substantially improves mAP over Faster R-CNN and several baselines under synthetically injected uniform random noise, with a modest gain on the original (naturally noisy) OPIXray dataset.

Significance. If the claims hold, the paper introduces a simple, plug-and-play augmentation direction for noisy-annotation object detection, which is a practical and under-explored problem for X-ray security imagery. The strengths are the breadth of experiments (two X-ray datasets plus MS-COCO, multiple noise rates, ablations of components, and comparison with several specialized baselines), the release of source code, and the clear presentation of the method. The main limitation is that the high-noise evidence is generated entirely from a uniform random synthetic noise protocol, whereas the paper's own characterization of real X-ray noise is structured (confusions among visually similar knife types). The significance is therefore real but currently contingent on the validity of the random-noise assumption.

major comments (3)
  1. [Section III-C, Eq. (1); Section IV-B; Section IV-C] The central probability argument, 1 - P_c^K, assumes that each same-label patch is independently corrupted with probability P_c under a random label-noise model. However, Section IV-B states that real category noise in OPIXray arises from the visual similarity between knife types (folding, utility, straight, multi-tool), i.e., a structured confusion pattern. Under such structured noise, K patches sharing the same noisy label may all be the same wrong class, so mixing does not increase the chance of including a true same-label item. All high-noise results (Tables I, II, IV) use the synthetic uniform random noise of Section IV-C; the only natural-noise evidence (Table VII) is at an estimated ~5% category noise, which cannot validate the high-noise claim. The authors should either test Mix-Paste under a structured noise model (e.g., a confusion matrix biased by visual similarity) or clearly scope the claims to random label noise.
  2. [Section III-C, Eqs. (1)-(3)] The claimed bounding-box benefit, 1 - P_b^K, is not supported by the algorithm as described. The mixed patch is pasted into the original (possibly noisy) bounding box, and the ground-truth box coordinates are unchanged, so mixing K patches does not by itself increase the probability that the box 'accurately bounds a correct prohibited item.' The empirical results in Fig. 5(b) suggest Mix-Paste helps under box-noise-only conditions, but the stated mechanism is not convincing. The theoretical explanation should be revised to describe how the augmentation actually interacts with box noise, and an ablation that isolates box noise (e.g., with clean labels) should be reported at the same noise rates as in the main tables.
  3. [Section III-D, Eq. (4)] LLS suppresses all non-background predictions whose label differs from the ground-truth label at high IoU, not only predictions corresponding to extra items introduced by Mix-Paste. In the very scenario addressed by the paper, the ground-truth label itself may be wrong; in that case the model's correct prediction of the true class would be suppressed, effectively training the model toward the noisy label. The paper demonstrates the benefit of LLS under random synthetic noise (Table XIV) but does not analyze this interaction with noisy ground-truth labels. A more selective suppression rule, or an experiment that separates LLS's effect on mixed-patch extra items from its effect on single-patch noisy labels, would strengthen the claim.
minor comments (5)
  1. [Eq. (2)] The definition of the edge-smoothing mask appears reversed: the formula assigns α ≈ 1 near the patch boundary (small d) and a constant λ in the interior, which would make the boundary dominated by the base patch rather than smoothing the edge. Please verify the formula or its verbal description; the ablation in Table VI suggests the mask has a minor effect, but the current definition is confusing.
  2. [Tables I-II] Several numbers in Table II have missing plus signs (e.g., '21.5 (1.6)' and '11.1(-2.2)'); the sign convention should be made consistent with Table I.
  3. [Conclusion] The conclusion refers to an 'LSS strategy'; the abbreviation should be 'LLS' for consistency with Section III-D.
  4. [References] The reference list appears twice in the arXiv version (pages 14-16); the final version should contain a single bibliography.
  5. [Table X] The complete collapse at p=1 (0.4% mAP) is a striking failure mode; a one- or two-sentence explanation beyond the train/test distribution mismatch would help readers understand the sensitivity to this hyperparameter.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the probabilistic motivation is a stated-assumption identity and the method's effectiveness is established empirically on held-out test sets.

full rationale

The claimed derivation in Section III-C, namely that mixing K patches with the same noisy label raises the probability of containing the correct item to 1 - Pc^K, is a conditional probability identity under the paper's explicitly stated random independent noise model. It is not a fitted quantity, and the paper does not treat this identity as sufficient evidence of efficacy. The actual support for Mix-Paste and the LLS strategy is empirical: Table I/II report held-out test performance under synthetic noise, Table VII shows gains on the original OPIXray dataset without added noise, and Table V ablates Mix-Paste and LLS independently. The synthetic noise injection in Section IV-C (Eq. 5) matches the independence assumptions of the motivation, which limits generalization to structured real noise such as visually similar knife categories, but matching a synthetic benchmark to a stated assumption is not circularity. Self-citations exist (e.g., [7] and [25] share authors with the present paper), but they appear only as background references and comparison baselines, not as load-bearing justification for Mix-Paste or LLS. No fitted parameter is renamed as a prediction, no uniqueness theorem is imported from prior work, and no ansatz is smuggled in via citation. The reported mAP gains are therefore not equivalent to the method's inputs by construction.

Assumptions & free parameters 5 free parameters · 3 assumptions · 0 invented entities

The method introduces no new physical entities. The main free parameters are augmentation hyperparameters (K, p, beta, delta) and an unspecified LLS IoU threshold. The central claims depend on assumptions that synthetic noise matches real noise and that linear blending yields discernible training items.

free parameters (5)
  • K (number of patches mixed) = 2
    Number of patches mixed in Eq. (1); selected by ablation on OPIXray test set (Fig. 3a), where K=4 and 5 degrade performance.
  • p (probability of applying Mix-Paste) = 0.6
    Chosen by ablation on OPIXray test set (Table X). At p=1.0, training collapses to mAP 0.4.
  • beta (edge smoothing region width) = 10%
    Set 'empirically' in Eq. (2) to control the smoothing area; no principled selection is given.
  • delta (bounding-box noise perturbation bound) = 0.3
    Uniform distribution bound in Eq. (5); also tested at 0.5, but not justified by real-world box errors.
  • LLS IoU threshold = not reported
    The threshold separating PB_neg, PB_fb, PB_pos, PB_pp in Section III-D is never given, yet it controls which losses are suppressed.
assumptions (3)
  • domain assumption Independent category noise across patches, so the probability of all K patches being wrong is Pc^K.
    Section III-C uses this independence to justify Mix-Paste; real mislabeling may be correlated by item similarity.
  • domain assumption The synthetic noise protocol in Eq. (5) is representative of real bounding-box and label errors.
    All main experiments use synthetic noise; only a small real-noise estimate (~5%) and a no-noise experiment are provided, so the transfer to naturally noisy data is unverified.
  • domain assumption Linear blending of item patches produces training signals from which a detector can learn the target item despite overlap.
    The probabilistic argument treats the blend as 'containing' the correct item if any source patch is correct; visually the blend is a superposition and may not be recognizable for a detector.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Augmentation Matters: A Mix-Paste Method for X-Ray Prohibited Item Detection under Noisy Annotations." pith.science (2026). https://pith.science/paper/VCCRAIBG

@misc{pith2026250101733,
  author       = {Pith},
  title        = {Pith review of: Augmentation Matters: A Mix-Paste Method for X-Ray Prohibited Item Detection under Noisy Annotations},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/VCCRAIBG}},
  note         = {Machine review of arXiv:2501.01733}
}
read the original abstract

Automatic X-ray prohibited item detection is vital for public safety. Existing deep learning-based methods all assume that the annotations of training X-ray images are correct. However, obtaining correct annotations is extremely hard if not impossible for large-scale X-ray images, where item overlapping is ubiquitous.As a result, X-ray images are easily contaminated with noisy annotations, leading to performance deterioration of existing methods.In this paper, we address the challenging problem of training a robust prohibited item detector under noisy annotations (including both category noise and bounding box noise) from a novel perspective of data augmentation, and propose an effective label-aware mixed patch paste augmentation method (Mix-Paste). Specifically, for each item patch, we mix several item patches with the same category label from different images and replace the original patch in the image with the mixed patch. In this way, the probability of containing the correct prohibited item within the generated image is increased. Meanwhile, the mixing process mimics item overlapping, enabling the model to learn the characteristics of X-ray images. Moreover, we design an item-based large-loss suppression (LLS) strategy to suppress the large losses corresponding to potentially positive predictions of additional items due to the mixing operation. We show the superiority of our method on X-ray datasets under noisy annotations. In addition, we evaluate our method on the noisy MS-COCO dataset to showcase its generalization ability. These results clearly indicate the great potential of data augmentation to handle noise annotations. The source code is released at https://github.com/wscds/Mix-Paste.

Figures

Figures reproduced from arXiv: 2501.01733 by the authors.

Figure 1
Figure 1. Examples in an X-ray dataset [1]. (a) Examples with noisy category [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Overview of our proposed method for training a robust prohibited item detector under noisy annotations. (a) illustrates Mix-Paste which mixes [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. (a) Ablation study results on the influence of the number of patches [PITH_FULL_IMAGE:figures/full_fig_p009_3.png] view at source ↗
Figures from the paper (3 more)
Figure 4
Figure 4. Figure 4: Examples of generated images on the OPIXray dataset. (a) The generated images with the linear combination. (b) The generated images with the [PITH_FULL_IMAGE:figures/full_fig_p011_4.png]
Figure 5
Figure 5. Figure 5: Performance comparison between Mix-Paste and the competing [PITH_FULL_IMAGE:figures/full_fig_p011_5.png]
Figure 6
Figure 6. Figure 6: Some detection results on the OPIXray dataset. (a) The detection results obtained by the baseline method (FRCNN). (b) The detection results obtained [PITH_FULL_IMAGE:figures/full_fig_p012_6.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

63 extracted references · 58 canonical work pages

  1. [1]

    Occluded prohibited items detection: An X-ray security inspection benchmark and de-occlusion attention module,

    Y . Wei, R. Tao, Z. Wu, Y . Ma, L. Zhang, and X. Liu, “Occluded prohibited items detection: An X-ray security inspection benchmark and de-occlusion attention module,” in Proc. ACM Int. Conf. Multimedia , 2020, pp. 138–146

  2. [2]

    SIXray: A large-scale security inspection X-ray benchmark for prohibited item discovery in overlapping images,

    C. Miao, L. Xie, F. Wan, C. Su, H. Liu, J. Jiao, and Q. Ye, “SIXray: A large-scale security inspection X-ray benchmark for prohibited item discovery in overlapping images,” in Proc. IEEE Conf. Comput. Vis. Pattern Recog., 2019, pp. 2114–2123

  3. [3]

    PIDray: A large-scale X-ray benchmark for real-world prohibited item detection,

    L. Zhang, L. Jiang, R. Ji, and H. Fan, “PIDray: A large-scale X-ray benchmark for real-world prohibited item detection,” in Proc. Int. J. Comput. Vis., 2023, pp. 3170–3192. 15

  4. [4]

    Towards real-world X-ray security inspection: A high-quality benchmark and lateral inhibition module for prohibited items detection,

    R. Tao, Y . Wei, X. Jiang, H. Li, H. Qin, J. Wang, Y . Ma, L. Zhang, and X. Liu, “Towards real-world X-ray security inspection: A high-quality benchmark and lateral inhibition module for prohibited items detection,” in Proc. Int. Conf. Comput. Vis. , 2021, pp. 10 923–10 932

  5. [5]

    ‘unex- pected item in the bagging area’: Anomaly detection in X-ray security images,

    L. D. Griffin, M. Caldwell, J. T. A. Andrews, and H. Bohler, “‘unex- pected item in the bagging area’: Anomaly detection in X-ray security images,” IEEE Trans. Inf. F orensics Security , pp. 1539–1553, 2019

  6. [6]

    Toward dual- view X-ray baggage inspection: A large-scale benchmark and adaptive hierarchical cross refinement for prohibited item discovery,

    B. Ma, T. Jia, M. Li, S. Wu, H. Wang, and D. Chen, “Toward dual- view X-ray baggage inspection: A large-scale benchmark and adaptive hierarchical cross refinement for prohibited item discovery,”IEEE Trans. Inf. F orensics Security, pp. 3866–3878, 2024

  7. [7]

    Dual- mode learning for multi-dataset X-ray security image detection,

    F. Yang, R. Jiang, Y . Yan, J.-H. Xue, B. Wang, and H. Wang, “Dual- mode learning for multi-dataset X-ray security image detection,” IEEE Trans. Inf. F orensics Security, pp. 3510–3524, 2024

  8. [8]

    DivideMix: Learning with noisy labels as semi-supervised learning,

    J. Li, R. Socher, and S. C. Hoi, “DivideMix: Learning with noisy labels as semi-supervised learning,” in Proc. Int. Conf. Learn. Represent., 2020

Show all 63 references
  1. [9]

    Co-teaching: Robust training of deep neural networks with extremely noisy labels,

    B. Han, Q. Yao, X. Yu, G. Niu, M. Xu, W. Hu, I. Tsang, and M. Sugiyama, “Co-teaching: Robust training of deep neural networks with extremely noisy labels,” inAdv. Neural Inform. Process. Syst., 2018, pp. 8536–8546

  2. [10]

    Symmetric cross entropy for robust learning with noisy labels,

    Y . Wang, X. Ma, Z. Chen, Y . Luo, J. Yi, and J. Bailey, “Symmetric cross entropy for robust learning with noisy labels,” in Proc. Int. Conf. Comput. Vis., 2019, pp. 322–330

  3. [11]

    Training object detectors with noisy data,

    S. Chadwick and P. Newman, “Training object detectors with noisy data,” in Proc. IEEE Intelligent V ehicles Symp. , 2019, pp. 1319–1325

  4. [12]

    Towards noise-resistant object detection with noisy annotations,

    J. Li, C. Xiong, R. Socher, and S. Hoi, “Towards noise-resistant object detection with noisy annotations,” arXiv preprint arXiv:2003.01285 , 2020

  5. [13]

    Learning with noisy class labels for instance segmentation,

    L. Yang, F. Meng, H. Li, Q. Wu, and Q. Cheng, “Learning with noisy class labels for instance segmentation,” in Proc. Eur . Conf. Comput. Vis., 2020, pp. 38–53

  6. [14]

    Robust object detection with inaccurate bounding boxes,

    C. Liu, K. Wang, H. Lu, Z. Cao, and Z. Zhang, “Robust object detection with inaccurate bounding boxes,” in Proc. Eur . Conf. Comput. Vis., 2022, pp. 53–69

  7. [15]

    Narrowing the gap: Improved detector training with noisy location annotations,

    S. Wang, J. Gao, B. Li, and W. Hu, “Narrowing the gap: Improved detector training with noisy location annotations,” IEEE Trans. Image Process., pp. 6369–6380, 2022

  8. [16]

    Understand- ing deep learning requires rethinking generalization,

    C. Zhang, S. Bengio, M. Hardt, B. Recht, and O. Vinyals, “Understand- ing deep learning requires rethinking generalization,” Commun. ACM , pp. 107–115, Feb 2017

  9. [17]

    A closer look at memorization in deep networks,

    D. Arpit, S. Jastrzebski, N. Ballas, D. Krueger, E. Bengio, M. S. Kanwal, T. Maharaj, A. Fischer, A. Courville, Y . Bengio et al. , “A closer look at memorization in deep networks,” in Proc. Int. Conf. Mach. Learn. , 2017, pp. 233–242

  10. [18]

    Microsoft COCO: Common objects in context,

    T.-Y . Lin, M. Maire, S. Belongie, J. Hays, P. Perona, D. Ramanan, P. Doll ´ar, and C. L. Zitnick, “Microsoft COCO: Common objects in context,” in Proc. Eur . Conf. Comput. Vis. , 2014, pp. 740–755

  11. [19]

    Detecting overlapped objects in X-ray security imagery by a label-aware mechanism,

    C. Zhao, L. Zhu, S. Dou, W. Deng, and L. Wang, “Detecting overlapped objects in X-ray security imagery by a label-aware mechanism,” IEEE Trans. Inf. F orensics Security, pp. 998–1009, 2022

  12. [20]

    Exploiting foreground and background separation for prohibited item detection in overlapping x-ray images,

    F. Shao, J. Liu, P. Wu, Z. Yang, and Z. Wu, “Exploiting foreground and background separation for prohibited item detection in overlapping x-ray images,” Pattern Recognition, p. 108261, 2022

  13. [21]

    Baggage threat recognition using deep low-rank broad learning detec- tor,

    D. Velayudhan, T. Hassan, A. H. Ahmed, E. Damiani, and N. Werghi, “Baggage threat recognition using deep low-rank broad learning detec- tor,” in IEEE Mediterranean Electrotechnical Conf. , 2022, pp. 966–971

  14. [22]

    Towards automatic threat detection: A survey of advances of deep learning within x-ray security imaging,

    S. Akcay and T. Breckon, “Towards automatic threat detection: A survey of advances of deep learning within x-ray security imaging,” Pattern Recognition, p. 108245, 2022

  15. [23]

    Computer vision on x-ray data in industrial production and security applications: A comprehensive survey,

    M. Rafiei, J. Raitoharju, and A. Iosifidis, “Computer vision on x-ray data in industrial production and security applications: A comprehensive survey,” IEEE Access , pp. 2445–2477, 2023

  16. [24]

    Recent advances in baggage threat detection: A comprehensive and systematic survey,

    D. Velayudhan, T. Hassan, E. Damiani, and N. Werghi, “Recent advances in baggage threat detection: A comprehensive and systematic survey,” ACM Computing Surveys , pp. 1–38, 2022

  17. [25]

    Gadet: A geometry-aware x-ray prohibited items detector,

    M. Li, B. Ma, H. Wang, D. Chen, and T. Jia, “Gadet: A geometry-aware x-ray prohibited items detector,” IEEE Sensors Journal , pp. 1665–1678, 2024

  18. [26]

    Improved regularization of convolutional neural networks with cutout,

    T. DeVries and G. W. Taylor, “Improved regularization of convolutional neural networks with cutout,” arXiv preprint arXiv:1708.04552 , 2017

  19. [27]

    mixup: Beyond empirical risk minimization,

    H. Zhang, M. Cisse, Y . N. Dauphin, and D. Lopez-Paz, “mixup: Beyond empirical risk minimization,” arXiv preprint arXiv:1710.09412 , 2017

  20. [28]

    Alignmixup: Improving representations by interpolating aligned features,

    S. Venkataramanan, E. Kijak, L. Amsaleg, and Y . Avrithis, “Alignmixup: Improving representations by interpolating aligned features,” in Proc. IEEE Conf. Comput. Vis. Pattern Recog. , 2022, pp. 19 174–19 183

  21. [29]

    Yolov4: Op- timal speed and accuracy of object detection,

    A. Bochkovskiy, C.-Y . Wang, and H.-Y . M. Liao, “Yolov4: Op- timal speed and accuracy of object detection,” arXiv preprint arXiv:2004.10934, 2020

  22. [30]

    Channel augmentation for visible- infrared re-identification,

    M. Ye, Z. Wu, C. Chen, and B. Du, “Channel augmentation for visible- infrared re-identification,” IEEE Trans. Pattern Anal. Mach. Intell. , pp. 2299–2315, 2024

  23. [31]

    Learning transferable visual models from natural language supervision,

    A. Radford, J. W. Kim, C. Hallacy, A. Ramesh, G. Goh, S. Agarwal, G. Sastry, A. Askell, P. Mishkin, J. Clark et al. , “Learning transferable visual models from natural language supervision,” in Proc. Int. Conf. Mach. Learn. , 2021, pp. 8748–8763

  24. [32]

    Adding conditional control to text-to-image diffusion models,

    L. Zhang, A. Rao, and M. Agrawala, “Adding conditional control to text-to-image diffusion models,” in Proc. Int. Conf. Comput. Vis. , 2023, pp. 3836–3847

  25. [33]

    Data augmentation for object detection via controllable diffusion models,

    H. Fang, B. Han, S. Zhang, S. Zhou, C. Hu, and W.-M. Ye, “Data augmentation for object detection via controllable diffusion models,” in Proc. IEEE Winter Conf. Appl. Comput. Vis. , 2024, pp. 1257–1266

  26. [34]

    Exploiting clip self-consistency to automate image augmentation for safety critical scenarios,

    S. S. Gannamaneni, F. Klein, M. Mock, and M. Akila, “Exploiting clip self-consistency to automate image augmentation for safety critical scenarios,” in Proc. IEEE Conf. Comput. Vis. Pattern Recog. , 2024, pp. 3594–3604

  27. [35]

    Op- erationalizing convolutional neural network architectures for prohibited object detection in x-ray imagery,

    T. W. Webb, N. Bhowmik, Y . F. A. Gaus, and T. P. Breckon, “Op- erationalizing convolutional neural network architectures for prohibited object detection in x-ray imagery,” inProc. IEEE Int. Conf. Mach. Learn. Appl., 2021, pp. 610–615

  28. [36]

    Generalized cross entropy loss for training deep neural networks with noisy labels,

    Z. Zhang and M. Sabuncu, “Generalized cross entropy loss for training deep neural networks with noisy labels,” in Adv. Neural Inform. Process. Syst., 2018, pp. 8792–8802

  29. [37]

    Co-learning: Learning from noisy labels with self-supervision,

    C. Tan, J. Xia, L. Wu, and S. Z. Li, “Co-learning: Learning from noisy labels with self-supervision,” in Proc. ACM Int. Conf. Multimedia , 2021, pp. 1405–1413

  30. [38]

    Training deep neural-networks using a noise adaptation layer,

    J. Goldberger and E. Ben-Reuven, “Training deep neural-networks using a noise adaptation layer,” in Proc. Int. Conf. Learn. Represent. , 2016, pp. 1–9

  31. [39]

    Making deep neural networks robust to label noise: A loss correction approach,

    G. Patrini, A. Rozza, A. Krishna Menon, R. Nock, and L. Qu, “Making deep neural networks robust to label noise: A loss correction approach,” in Proc. IEEE Conf. Comput. Vis. Pattern Recog. , 2017, pp. 1944–1952

  32. [40]

    Part-dependent label noise: Towards instance-dependent label noise,

    X. Xia, T. Liu, B. Han, N. Wang, M. Gong, H. Liu, G. Niu, D. Tao, and M. Sugiyama, “Part-dependent label noise: Towards instance-dependent label noise,” in Adv. Neural Inform. Process. Syst., 2020, pp. 7597–7610

  33. [41]

    Provably end-to-end label-noise learning without anchor points,

    X. Li, T. Liu, B. Han, G. Niu, and M. Sugiyama, “Provably end-to-end label-noise learning without anchor points,” in Proc. Int. Conf. Mach. Learn., 2021, pp. 6403–6413

  34. [42]

    Robust loss functions under label noise for deep neural networks,

    A. Ghosh, H. Kumar, and P. S. Sastry, “Robust loss functions under label noise for deep neural networks,” in Proc. AAAI Conf. Artif. Intell. , 2017, pp. 1919–1925

  35. [43]

    Normalized loss functions for deep learning with noisy labels,

    X. Ma, H. Huang, Y . Wang, S. Romano, S. Erfani, and J. Bailey, “Normalized loss functions for deep learning with noisy labels,” in Proc. Int. Conf. Mach. Learn. , 2020, pp. 6543–6553

  36. [44]

    MentorNet: Learning data-driven curriculum for very deep neural networks on corrupted labels,

    L. Jiang, Z. Zhou, T. Leung, L.-J. Li, and L. Fei-Fei, “MentorNet: Learning data-driven curriculum for very deep neural networks on corrupted labels,” in Proc. Int. Conf. Mach. Learn. , 2018, pp. 2304– 2313

  37. [45]

    How does disagreement help generalization against label corruption?

    X. Yu, B. Han, J. Yao, G. Niu, I. Tsang, and M. Sugiyama, “How does disagreement help generalization against label corruption?” in Proc. Int. Conf. Mach. Learn. , 2019, pp. 7164–7173

  38. [46]

    Combating noisy labels by agreement: A joint training method with co-regularization,

    H. Wei, L. Feng, X. Chen, and B. An, “Combating noisy labels by agreement: A joint training method with co-regularization,” in Proc. IEEE Conf. Comput. Vis. Pattern Recog. , 2020, pp. 13 726–13 735

  39. [47]

    Decoupling representation and classifier for long-tailed recognition,

    B. Kang, S. Xie, M. Rohrbach, Z. Yan, A. Gordo, J. Feng, and Y . Kalantidis, “Decoupling representation and classifier for long-tailed recognition,” arXiv preprint arXiv:1910.09217 , 2020

  40. [48]

    PurifyNet: A robust person re-identification model with noisy labels,

    M. Ye and P. C. Yuen, “PurifyNet: A robust person re-identification model with noisy labels,” IEEE Trans. Inf. F orensics Security, pp. 2655– 2666, 2020

  41. [49]

    Collaborative refining for person re-identification with label noise,

    M. Ye, H. Li, B. Du, J. Shen, L. Shao, and S. C. H. Hoi, “Collaborative refining for person re-identification with label noise,” IEEE Trans. Image Process., pp. 379–391, 2022

  42. [50]

    Structure-aware positional transformer for visible-infrared person re-identification,

    C. Chen, M. Ye, M. Qi, J. Wu, J. Jiang, and C.-W. Lin, “Structure-aware positional transformer for visible-infrared person re-identification,” IEEE Trans. Image Process. , pp. 2352–2364, 2022

  43. [51]

    The good, the bad and the ugly: Evaluating convolutional neural networks for prohibited item detection using real and synthetically composited x-ray imagery,

    N. Bhowmik, Q. Wang, Y . F. A. Gaus, M. Szarek, and T. P. Breckon, “The good, the bad and the ugly: Evaluating convolutional neural networks for prohibited item detection using real and synthetically composited x-ray imagery,” arXiv preprint arXiv:1909.11508 , 2019

  44. [52]

    Rwsc-fusion: Region-wise style-controlled fusion network for the prohibited x-ray 16 security image synthesis,

    L. Duan, M. Wu, L. Mao, J. Yin, J. Xiong, and X. Li, “Rwsc-fusion: Region-wise style-controlled fusion network for the prohibited x-ray 16 security image synthesis,” in Proc. IEEE Conf. Comput. Vis. Pattern Recog., 2023, pp. 22 398–22 407

  45. [53]

    A logarithmic x-ray imaging model for baggage inspection: Simulation and object detection,

    D. Mery and A. K. Katsaggelos, “A logarithmic x-ray imaging model for baggage inspection: Simulation and object detection,” in IEEE Conf. Comput. Vis. Pattern Recog. Worksh. , 2017, pp. 57–65

  46. [54]

    Threat image projection (tip) into x-ray images of cargo containers for training humans and machines,

    T. W. Rogers, N. Jaccard, E. D. Protonotarios, J. Ollier, E. J. Morton, and L. D. Griffin, “Threat image projection (tip) into x-ray images of cargo containers for training humans and machines,” in IEEE Int. Carnahan Conf. Security Technol. , 2016, pp. 1–7

  47. [55]

    CutMix: Reg- ularization strategy to train strong classifiers with localizable features,

    S. Yun, D. Han, S. J. Oh, S. Chun, J. Choe, and Y . Yoo, “CutMix: Reg- ularization strategy to train strong classifiers with localizable features,” in Proc. Int. Conf. Comput. Vis. , 2019, pp. 6023–6032

  48. [56]

    Saliencymix: A saliency guided data augmentation strategy for better regularization,

    A. Uddin, M. Monira, W. Shin, T. Chung, S.-H. Bae et al., “Saliencymix: A saliency guided data augmentation strategy for better regularization,” in Proc. Int. Conf. Learn. Represent. , 2021

  49. [57]

    Attentive cutmix: An enhanced data augmentation approach for deep learning based image classification,

    D. Walawalkar, Z. Shen, Z. Liu, and M. Savvides, “Attentive cutmix: An enhanced data augmentation approach for deep learning based image classification,” in ICASSP, 2020

  50. [58]

    Faster R-CNN: Towards real-time object detection with region proposal networks,

    S. Ren, K. He, R. Girshick, and J. Sun, “Faster R-CNN: Towards real-time object detection with region proposal networks,” IEEE Trans. Pattern Anal. Mach. Intell. , pp. 1137–1149, 2017

  51. [59]

    ImageNet classification with deep convolutional neural networks,

    A. Krizhevsky, I. Sutskever, and G. E. Hinton, “ImageNet classification with deep convolutional neural networks,” Commun. ACM , pp. 84–90, 2017

  52. [60]

    Focal loss for dense object detection,

    T.-Y . Lin, P. Goyal, R. Girshick, K. He, and P. Doll ´ar, “Focal loss for dense object detection,” in Proc. Int. Conf. Comput. Vis., 2017, pp. 2980– 2988

  53. [61]

    Cascade R-CNN: Delving into high quality object detection,

    Z. Cai and N. Vasconcelos, “Cascade R-CNN: Delving into high quality object detection,” in Proc. IEEE Conf. Comput. Vis. Pattern Recog. , 2018, pp. 6154–6162

  54. [62]

    Bridging the gap between anchor-based and anchor-free detection via adaptive training sample selection,

    S. Zhang, C. Chi, Y . Yao, Z. Lei, and S. Z. Li, “Bridging the gap between anchor-based and anchor-free detection via adaptive training sample selection,” in Proc. IEEE Conf. Comput. Vis. Pattern Recog. , 2020, pp. 9759–9768

  55. [63]

    Mmdetection: Open mmlab detection toolbox and benchmark,

    K. Chen, J. Wang, J. Pang, Y . Cao, Y . Xiong, X. Li, S. Sun, W. Feng, Z. Liu, J. Xu et al. , “Mmdetection: Open mmlab detection toolbox and benchmark,” arXiv preprint arXiv:1906.07155 , 2019

Pith tools

Reviewed August 10, 2026 · model on record in the stance chip above.