Pith. sign in

REVIEW 4 major objections 5 minor 46 references

An approach based on class activation maps for investigating the effects of data augmentation on neural networks for image classification

T0 review · 4 major / 5 minor · reviewed 2026-08-15 · deepseek-v4-flash

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

Pith's one-line read This paper claims that data augmentation's effects on a CNN's learned attention can be measured automatically by comparing class activation maps with similarity metrics, and that correlation analysis of those metrics surfaces distinct…

desk verdict Careful, honest experimental study of a CAM-comparison methodology for data augmentation effects, but the 'impact profiles' claim is statistically unsupported and the key frequency table is internally inconsistent. read the letter →

arxiv 2505.12581 v1 pith:42SH6Q34 submitted 2025-05-19 cs.LG cs.AIcs.CV

Lucas M. Dorneles , Luan Fonseca Garcia , Joel Luís Carbonera This is my paper · ORCID
classification cs.LGcs.AIcs.CV
keywords dataaugmentationclassactivationmapsGrad-CAMexplainabilityimageclassificationconvolutionalneuralnetworksmodelinterpretabilityCAMsimilaritymetrics
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 claims that the effects of data augmentation on a trained image classifier can be measured automatically, quantitatively, and at scale by comparing class activation maps (CAMs): heatmaps of which pixels each model considers important for its prediction. It proposes a generic pipeline: train a baseline model and several models per augmentation on identically ordered data, generate a Grad-CAM for every test image using the predicted class as target, compute six pixel-level and distributional similarity metrics between each augmented model's maps and the baseline's, then analyze the metric distributions and how the metrics correlate across augmentations. Applied to CIFAR-10 with EfficientNet B0 and seven augmentation techniques, the pipeline finds that augmented models share broad attention regions with the baseline but diverge on the most discriminative pixels, and that repeated counts of most- and least-correlated augmentation pairs suggest distinct behavioral clusters. If correct, the value is that augmentation effects no longer depend on a human eyeballing heatmap pairs or on accuracy numbers alone.

What carries the argument

The central object is the Grad-CAM class activation map: a heatmap, normalized to $[0,1]$, taken at the last convolutional layer, that scores each pixel's contribution to the model's predicted class. The mechanism that carries the argument is the metric set computed between each augmented model's maps and the baseline's -- MAD and MSD for magnitude of difference, Pearson and Spearman for linear and monotonic co-variation, Overlap Rate for agreement on the top-$Y\%$ most important pixels, and Class-KLD for divergence of the prediction distributions -- followed by aggregation across multiple training seeds per augmentation. The load-bearing analytical move is the relative correlation analysis: instead of reading absolute metric values, the paper counts how often each augmentation pair appears among the most- and least-correlated pairs across all metrics, and those counts are what surface the claimed behavioral clusters.

What would settle it

Run the same pipeline between two baseline models trained from different random initial states with no augmentation in either; if the MAD/MSD distributions and the cross-metric correlation counts match those of augmented-versus-baseline comparisons, the metrics are capturing general training variability rather than augmentation effects. As a second check, regenerate all CAMs with the ground-truth label as the target class instead of the predicted class and recompute the correlation counts: if the Cutmix-Affine-Color Jitter cluster dissolves, the claimed impact profiles are an artifact of prediction disagreement rather than of learned attention.

Watch

Extended reading notes

Core claim

The central claim is that Grad-CAM maps, compared pairwise between a baseline model and models trained with data augmentation, carry measurable and meaningful information about how augmentation changes what a CNN learns to look at. Concretely, the paper argues that (1) a set of six similarity metrics applied to CAMs -- mean absolute and squared difference, Pearson and Spearman correlation, top-$Y\%$ overlap rate, and KL divergence between class-prediction distributions -- produces stable summary statistics over an entire test set; (2) these statistics reveal systematic effects, such as overlap rate falling as the activation threshold tightens, meaning augmented models agree with the baseline on broad attention regions but not on the sharpest discriminative pixels, with the cross-augmentation correlations of that metric also weakening at stricter thresholds; and (3) counting which augmentation pairs rank as most and least correlated across all metrics suggests recurring impact profiles, clustering Cutmix, Affine, and Color Jitter together, pairing Gaussian Blur with Elastic Transform, and leaving Equalization as its own profile. The authors present this as a first, extensible methodology for quantitative augmentation-impact analysis, and they are candid that individual augmentation methods remain hard to tell apart from absolute metric values alone.

Load-bearing premise

The load-bearing premise is that pixel-level similarity or correlation between two Grad-CAM heatmaps faithfully measures how differently two models actually classify an image, even when the two models predict different classes for that image; if CAM similarity does not track behavioral difference, every metric in the pipeline loses its meaning.

Editorial extensions

If this is right

  • Augmentation effects can be audited over an entire test set with no human annotation of important regions, closing the scalability gap that limited qualitative CAM comparisons.
  • The methodology is generic across datasets, architectures, CAM generators, and metric choices, so the same 13 steps can be reused for other augmentation families or adapted to other sources of behavioral change in a classifier.
  • Overlap rate at stricter thresholds (top-$Y\%$ with small $Y$) is the metric most sensitive to augmentation, so studies that use only magnitude-based differences such as MAD or MSD are likely to miss the sharpest behavioral divergences.
  • The correlation counts imply augmentation techniques come in behavioral clusters -- geometric and mixing transforms (Cutmix, Affine, Color Jitter) versus smoothing transforms (Gaussian Blur, Elastic) -- which, if stable, would let practitioners choose augmentations by expected behavioral profile rather than by accuracy alone.
  • Because Class-KLD stays small while CAM differences are large, the final classification layers appear to absorb much of the behavioral shift, suggesting that augmentation's detectable footprint is strongest before the classifier head and may be invisible to prediction-based analyses.

Reading between the lines

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

  • A control the paper does not run would directly test the profiling claim: compare CAMs between two baselines trained from different random initial states, with no augmentation in either. If those metric distributions and correlation counts matched the augmented-versus-baseline ones, the measured effects would be dominated by training variability rather than by augmentation.
  • The clustering result yields a testable prediction the paper leaves implicit: augmentations inside the same cluster should be roughly interchangeable when behavioral similarity is the goal, while augmentations from different clusters should produce the most divergent footprints when combined -- an experiment one could run by applying two cross-cluster augmentations jointly and checking whether the
  • Because the pipeline needs only a target class, the last convolutional layer, and a test set, it could be pointed at other model interventions such as fine-tuning, pruning, or adversarial retraining, turning it from an augmentation profiler into a general behavioral-difference tool.
  • The predicted-versus-ground-truth target choice remains an under-explored degree of freedom: the paper chose predicted classes and showed prediction-disagreement effects are minimal, but a perfectly calibrated or adversarially robust model might show a different footprint, which is a cheap follow-up experiment with the same code path.
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

4 major / 5 minor

Summary. The paper proposes a generic, scalable methodology for quantitatively analyzing how data augmentation techniques change the behavior of a CNN image classifier. The method generates Grad-CAM maps for a baseline model and for models trained on augmented data, computes eight similarity/difference metrics between baseline and augmented maps over the test set, and then analyzes those metrics through boxplots and cross-augmentation correlation maps. The method is instantiated on CIFAR-10 with EfficientNet-B0, seven augmentation techniques, and three starting seeds per technique, with detailed controls for batch ordering and random state. The central empirical claim, stated in Section 1 and developed in Section 5.2.6, is that relative frequencies of strongly/weakly correlated augmentation pairs suggest the existence of distinct impact profiles (clusters) among augmentation techniques.

Significance. If the central claim were established, the methodology would be a useful annotation-free tool for comparing data augmentation effects at scale, complementing performance-based comparisons. The paper's strengths are its clearly described experimental protocol, the use of multiple seeds, explicit controls for randomization and data ordering, and the robustness check in Appendix C. These reproducibility-oriented choices are commendable. However, the key empirical payload—the existence of distinct impact profiles—currently rests on frequency counts derived from correlations that are not tested against a null model and that are computed over a common set of test images without controlling for shared image-level factors. The methodology as a proposal is viable, but the evidence for the cluster interpretation is not yet convincing.

major comments (4)
  1. [Section 5.2.6, Tables 1 and 2] The cluster interpretation is not supported by the statistical evidence presented. The counts in Tables 1 and 2 are computed from pairwise correlations across augmentations over the same 10,000 test images (Section 4, steps 11-12). Any image-level factor that influences CAM divergence for all augmentations—such as image difficulty or ambiguity—will inflate all pairwise correlations, producing uniformly moderate values like the reported MAD range of 0.58-0.67 and the general pattern of correlations above 0.5. The manuscript itself notes that absolute correlation values are 'not significantly different across pairs' (Section 5.2.6), which is exactly why the relative ranking used for Tables 1 and 2 requires a null model, permutation test, or partial-correlation analysis that removes the common image-level component. Without such a control, the frequency counts cannot be interpreted as evidence of distinct impact profiles rather than as an artifact of a shared confounding factor.
  2. [Table 1] Table 1 is internally inconsistent. The pair 'Gaussian Blur - Elastic Transform' appears twice, once with count 5 and once with count 1, and the text in Section 5.2.6 states that this pair appears '7 out of 8 times,' which does not match either row or their sum. The duplicate row also breaks the table's total-slot accounting. Because Tables 1 and 2 are the primary evidence for the claimed profiles, this error undermines the reliability of the central result and must be corrected or explained.
  3. [Section 5.2.6, metric dependence] The eight metrics are not independent evidence, and the '8 out of 8 times' statements in Section 5.2.6 and Table 2 overstate the support. Three of the metrics are overlap-rate variants at Y=20, 10, and 5, which are near-duplicate transformations of the same underlying information, and MAD/MSD similarly form a correlated pair. If the metric families are counted independently, the effective number of independent tests is much smaller than eight. The analysis should either aggregate the counts over metric families or explicitly account for the dependence among metrics before making claims about consistency across 'all' metrics.
  4. [Section 4 steps 10-11, Appendix C] The comparability of CAMs generated for different predicted classes remains a load-bearing modeling assumption. Since the target class for Grad-CAM is the model's own prediction, whenever the baseline and augmented models disagree on the predicted class, the two CAMs encode importance for different classes, and the pixel-level metrics mix class-specific attention with augmentation-induced changes. The Appendix C check concludes that the impact is minimal, but it is based on aggregate boxplot distributions for the overlap-rate metric only, not on the correlation maps or the frequency tables that support the profile claim. The comparability assumption therefore needs to be validated for the specific quantities used in the main analysis.
minor comments (5)
  1. [Section 5.1, Class-KLD definition] The regularization constant epsilon in the Class-KLD equation is never given a value, and the formula KLD(P,Q) = sum_i Q_i log(epsilon + Q_i/P_i) is not a standard KLD. Please state the numerical value of epsilon and justify how this regularized version behaves as a divergence.
  2. [Appendix B] Reproducibility would be improved by reporting the actual seed values used for the starting states and for the dataset shuffle, rather than referring only to 'a specific, separate, stable seed.'
  3. [Section 1 and Section 6] The abstract and introduction state that the experiments suggest 'different impact profiles' over trained models, but Section 6 correctly cautions that the absolute differences in correlation values are small. Please align the wording in the abstract and introduction with the more cautious conclusion.
  4. [Section 5.2.2] In the discussion of the Pearson and Spearman boxplots, the phrase 'inverse correlation' is used without reporting the proportion of test images for which the correlation is negative. A quantitative statement would make the observation more precise.
  5. [Figure 13] The training-time comparison figure would be clearer if the y-axis label and units were specified.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the CAM metrics are direct measurements, not fitted predictions, and no load-bearing step reduces to its own inputs.

full rationale

The paper's pipeline is an observational measurement exercise: Grad-CAM maps are generated for baseline and augmented models (Section 4, steps 10-11), eight comparison metrics are computed per image (MAD, MSD, Pearson, Spearman, three overlap-rate thresholds, Class-KLD), and the resulting distributions and pairwise between-augmentation correlations are inspected (Section 5.2). No parameter is fitted to any subset of the data and then renamed as a prediction; the 'impact profiles' are a post hoc reading of correlation patterns, not quantities derived from the conclusion by construction. The methodology is self-contained in the sense that the metrics are defined independently of the claimed clusters, and the cluster interpretation is openly tentative ('supposing the existence of behavior profiles', Section 5.2.6), with the paper itself noting the absolute correlation differences are small and that it 'struggled to investigate the effects of data augmentation on an individual level' (Section 6). Self-citations appear only as routine prior-work references in the introduction and are not load-bearing. The internal inconsistency in Table 1 (the duplicated 'Gaussian Blur - Elastic Transform' row) and the absence of significance tests are correctness or robustness concerns, not circularity: no equation or definition equates the output with an input, and no cited result is invoked to forbid alternative interpretations. The analysis therefore does not exhibit any of the enumerated circularity patterns.

Assumptions & free parameters 3 free parameters · 4 assumptions · 0 invented entities

The paper introduces no new physical or conceptual entities. The main assumptions are about the validity of Grad-CAM as a behavioral probe and the comparability of maps across models with potentially different predictions. Free parameters are limited to manually chosen thresholds and seeds.

free parameters (3)
  • Overlap Rate threshold Y = 20, 10, 5
    Hand-selected thresholds for defining important pixel regions; results depend on this choice, and no sensitivity analysis is provided.
  • Class-KLD regularization epsilon = not specified numerically
    Introduced to avoid log of zero in KL divergence; no sensitivity analysis is provided.
  • Number of starting states |X| = 3
    Arbitrary choice; more seeds would improve statistical robustness, and no power analysis is given.
assumptions (4)
  • domain assumption Grad-CAM maps reflect the discriminative image regions used by the model for classification.
    The entire comparison relies on this assumption; if Grad-CAM maps are unreliable, the metrics measure noise. Invoked in Section 2.4 and Section 4 step 10.
  • domain assumption Pixel-level similarity between CAMs captures meaningful differences in learned behavior.
    The metrics compare CAMs pointwise or via correlation; this assumes spatial correspondence is meaningful, which is questionable for augmentations like cropping or elastic transforms. Invoked in Section 4 steps 10-12.
  • domain assumption Using the predicted class as the Grad-CAM target yields comparable maps.
    When models disagree, CAMs explain different classes; the authors test this in Appendix C but still rely on this choice for the main analysis. Invoked in Section 5.1.
  • domain assumption Seeding and batch-order controls isolate the effects of augmentation.
    The paper assumes that after controlling for seeds and image order, observed differences are solely due to augmentation. This is a reasonable but untestable assumption. Invoked in Appendix B.

how reviews work

0 comments
Cite this review

Pith. "Pith review of An approach based on class activation maps for investigating the effects of data augmentation on neural networks for image classification." pith.science (2026). https://pith.science/paper/42SH6Q34

@misc{pith2026250512581,
  author       = {Pith},
  title        = {Pith review of: An approach based on class activation maps for investigating the effects of data augmentation on neural networks for image classification},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/42SH6Q34}},
  note         = {Machine review of arXiv:2505.12581}
}
read the original abstract

Neural networks have become increasingly popular in the last few years as an effective tool for the task of image classification due to the impressive performance they have achieved on this task. In image classification tasks, it is common to use data augmentation strategies to increase the robustness of trained networks to changes in the input images and to avoid overfitting. Although data augmentation is a widely adopted technique, the literature lacks a body of research analyzing the effects data augmentation methods have on the patterns learned by neural network models working on complex datasets. The primary objective of this work is to propose a methodology and set of metrics that may allow a quantitative approach to analyzing the effects of data augmentation in convolutional networks applied to image classification. An important tool used in the proposed approach lies in the concept of class activation maps for said models, which allow us to identify and measure the importance these models assign to each individual pixel in an image when executing the classification task. From these maps, we may then extract metrics over the similarities and differences between maps generated by these models trained on a given dataset with different data augmentation strategies. Experiments made using this methodology suggest that the effects of these data augmentation techniques not only can be analyzed in this way but also allow us to identify different impact profiles over the trained models.

Figures

Figures reproduced from arXiv: 2505.12581 by the authors.

Figure 1
Figure 1. Example of affine transformation augmentation. [PITH_FULL_IMAGE:figures/full_fig_p004_1.png] view at source ↗
Figure 3
Figure 3. Example of color jitter augmentation. • Elastic transforms, a transform that works based on displacement vectors applied to all pixels. An example is shown in [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗
Figure 4
Figure 4. Example of elastic transformation augmentation. [PITH_FULL_IMAGE:figures/full_fig_p005_4.png] view at source ↗
Figures from the paper (16 more)
Figure 6
Figure 6. Figure 6: Example of Gaussian blur augmentation. • Random Cropping, which crops the input image at a random location and resizes the cropped area to be the same size as the input image’s. An example is shown in [PITH_FULL_IMAGE:figures/full_fig_p006_6.png]
Figure 7
Figure 7. Figure 7: Example of random cropping augmentation. [PITH_FULL_IMAGE:figures/full_fig_p006_7.png]
Figure 9
Figure 9. Figure 9: Evolution of the accuracy on the test set across 30 training epochs for the three candidate architectures. [PITH_FULL_IMAGE:figures/full_fig_p010_9.png]
Figure 11
Figure 11. Figure 11: Evolution of the recall on the test set across 30 training epochs for the three candidate architectures. [PITH_FULL_IMAGE:figures/full_fig_p011_11.png]
Figure 12
Figure 12. Figure 12: Evolution of the F1-score on the test set across 30 training epochs for the three candidate architectures. [PITH_FULL_IMAGE:figures/full_fig_p011_12.png]
Figure 14
Figure 14. Figure 14: Bar plot of the accuracy performance metric for all models over the test image set. [PITH_FULL_IMAGE:figures/full_fig_p015_14.png]
Figure 15
Figure 15. Figure 15: Bar plot of the precision performance metric for all models over the test image set. [PITH_FULL_IMAGE:figures/full_fig_p015_15.png]
Figure 17
Figure 17. Figure 17: Bar plot of the F1-score performance metric for all models over the test image set. [PITH_FULL_IMAGE:figures/full_fig_p016_17.png]
Figure 19
Figure 19. Figure 19: Correlation map for overlap rate with threshold [PITH_FULL_IMAGE:figures/full_fig_p017_19.png]
Figure 21
Figure 21. Figure 21: Correlation map for overlap rate with threshold [PITH_FULL_IMAGE:figures/full_fig_p018_21.png]
Figure 23
Figure 23. Figure 23: Correlation map for overlap rate with threshold [PITH_FULL_IMAGE:figures/full_fig_p019_23.png]
Figure 24
Figure 24. Figure 24: Boxplot for Pearson Correlation Coefficient, showing the distribution of the metric values over the test set [PITH_FULL_IMAGE:figures/full_fig_p020_24.png]
Figure 27
Figure 27. Figure 27: Correlation map for Spearman Correlation Coefficient, showing the correlation of the metric between [PITH_FULL_IMAGE:figures/full_fig_p021_27.png]
Figure 29
Figure 29. Figure 29: Correlation map for Mean Absolute Difference, showing the correlation of the metric between augmentations [PITH_FULL_IMAGE:figures/full_fig_p023_29.png]
Figure 31
Figure 31. Figure 31: Correlation map for Mean Squared Difference, showing the correlation of the metric between augmentations [PITH_FULL_IMAGE:figures/full_fig_p024_31.png]
Figure 32
Figure 32. Figure 32: Boxplot for Class-KLD, showing the distribution of the metric values over the test set for each augmentation. [PITH_FULL_IMAGE:figures/full_fig_p025_32.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

46 extracted references · 25 canonical work pages

  1. [1]

    Multiscale context features for geological image classification

    Matheus V Todescato, Luan Fonseca Garcia, Dennis Giovani Balreira, and Joel Lu \' s Carbonera. Multiscale context features for geological image classification. In ICEIS (1), pages 407--418, 2023

    work page 2023

  2. [2]

    Multiscale patch-based feature graphs for image classification

    Matheus V Todescato, Luan F Garcia, Dennis G Balreira, and Joel L Carbonera. Multiscale patch-based feature graphs for image classification. Expert Systems with Applications, 235: 0 121116, 2024

    work page 2024

  3. [3]

    Todescato, and Joel Lu \' s Carbonera

    Erick da Silva Puls, Matheus V. Todescato, and Joel Lu \' s Carbonera. An evaluation of pre-trained models for feature extraction in image classification. In ICEIS, pages 123--135, 2024

    work page 2024

  4. [4]

    Analyzing Effects of Mixed Sample Data Augmentation on Model Interpretability

    Soyoun Won, Sung-Ho Bae, and Seong Tae Kim. Analyzing effects of mixed sample data augmentation on model interpretability. arXiv preprint arXiv:2303.14608, 2023

    work page Pith review arXiv 2023

  5. [5]

    A survey on image data augmentation for deep learning

    Connor Shorten and Taghi M Khoshgoftaar. A survey on image data augmentation for deep learning. Journal of big data, 6 0 (1): 0 1--48, 2019

    work page 2019

  6. [6]

    Towards explainable deep neural networks (xdnn)

    Plamen Angelov and Eduardo Soares. Towards explainable deep neural networks (xdnn). Neural Networks, 130: 0 185--194, 2020

    work page 2020

  7. [7]

    A survey on neural network interpretability

    Yu Zhang, Peter Ti n o, Ale s Leonardis, and Ke Tang. A survey on neural network interpretability. IEEE Transactions on Emerging Topics in Computational Intelligence, 5 0 (5): 0 726--742, 2021

    work page 2021

  8. [8]

    Interpretable deep convolutional neural networks via meta-learning

    Xuan Liu, Xiaoguang Wang, and Stan Matwin. Interpretable deep convolutional neural networks via meta-learning. In 2018 International Joint Conference on Neural Networks (IJCNN), pages 1--9. IEEE, 2018

    work page 2018

Show all 46 references
  1. [9]

    An analysis of explainability methods for convolutional neural networks

    Lynn Vonder Haar, Timothy Elvira, and Omar Ochoa. An analysis of explainability methods for convolutional neural networks. Engineering Applications of Artificial Intelligence, 117: 0 105606, 2023

    2023

  2. [10]

    On interpretability of artificial neural networks: A survey

    Feng-Lei Fan, Jinjun Xiong, Mengzhou Li, and Ge Wang. On interpretability of artificial neural networks: A survey. IEEE Transactions on Radiation and Plasma Medical Sciences, 5 0 (6): 0 741--760, 2021

    2021

  3. [11]

    Eigen-cam: Class activation map using principal components

    Mohammed Bany Muhammad and Mohammed Yeasin. Eigen-cam: Class activation map using principal components. In 2020 international joint conference on neural networks (IJCNN), pages 1--7. IEEE, 2020

    2020

  4. [12]

    The building blocks of interpretability

    Chris Olah, Arvind Satyanarayan, Ian Johnson, Shan Carter, Ludwig Schubert, Katherine Ye, and Alexander Mordvintsev. The building blocks of interpretability. Distill, 3 0 (3): 0 e10, 2018

    2018

  5. [13]

    Explaining the effect of data augmentation on image classification tasks, 2020

    Jerry Tang, Manasi Sharma, and Ruohan Zhang. Explaining the effect of data augmentation on image classification tasks, 2020

    2020

  6. [14]

    On the impact of interpretability methods in active image augmentation method

    Fl \'a vio Arthur Oliveira Santos, Cleber Zanchettin, Leonardo Nogueira Matos, and Paulo Novais. On the impact of interpretability methods in active image augmentation method. Logic Journal of the IGPL, 30 0 (4): 0 611--621, 2022

    2022

  7. [15]

    Comparing data augmentation strategies for deep image classification

    Sarah O'Gara and Kevin McGuinness. Comparing data augmentation strategies for deep image classification. Technical report, Technological University Dublin, 2019

    2019

  8. [16]

    Comparison of different image data augmentation approaches

    Loris Nanni, Michelangelo Paci, Sheryl Brahnam, and Alessandra Lumini. Comparison of different image data augmentation approaches. Journal of imaging, 7 0 (12): 0 254, 2021

    2021

  9. [17]

    The effectiveness of data augmentation in image classification using deep learning

    Luis Perez and Jason Wang. The effectiveness of data augmentation in image classification using deep learning. arXiv preprint arXiv:1712.04621, 2017

    2017 arXiv

  10. [18]

    Data augmentation for hyperspectral image classification with deep cnn

    Wei Li, Chen Chen, Mengmeng Zhang, Hengchao Li, and Qian Du. Data augmentation for hyperspectral image classification with deep cnn. IEEE Geoscience and Remote Sensing Letters, 16 0 (4): 0 593--597, 2018

    2018

  11. [19]

    Gridmask data augmentation

    Pengguang Chen, Shu Liu, Hengshuang Zhao, and Jiaya Jia. Gridmask data augmentation. arXiv preprint arXiv:2001.04086, 2020

    2001 arXiv

  12. [20]

    A survey of mix-based data augmentation: Taxonomy, methods, applications, and explainability

    Chengtai Cao, Fan Zhou, Yurou Dai, and Jianping Wang. A survey of mix-based data augmentation: Taxonomy, methods, applications, and explainability. arXiv preprint arXiv:2212.10888, 2022

    2022 arXiv

  13. [21]

    Saliencymix: A saliency guided data augmentation strategy for better regularization

    AFM Uddin, Mst Monira, Wheemyung Shin, TaeChoong Chung, Sung-Ho Bae, et al. Saliencymix: A saliency guided data augmentation strategy for better regularization. arXiv preprint arXiv:2006.01791, 2020

    2006 arXiv

  14. [22]

    Grad-cam: Visual explanations from deep networks via gradient-based localization

    Ramprasaath R Selvaraju, Michael Cogswell, Abhishek Das, Ramakrishna Vedantam, Devi Parikh, and Dhruv Batra. Grad-cam: Visual explanations from deep networks via gradient-based localization. In Proceedings of the IEEE international conference on computer vision, pages 618--626, 2017

    2017

  15. [23]

    Learning multiple layers of features from tiny images

    Alex Krizhevsky, Geoffrey Hinton, et al. Learning multiple layers of features from tiny images. Technical report, University of Toronto, 2009

    2009

  16. [24]

    Efficientnet: Rethinking model scaling for convolutional neural networks

    Mingxing Tan and Quoc Le. Efficientnet: Rethinking model scaling for convolutional neural networks. In International conference on machine learning, pages 6105--6114. PMLR, 2019

    2019

  17. [25]

    Introduction to machine learning

    Tom Mitchell. Introduction to machine learning. Machine learning, 7: 0 2--5, 1997

    1997

  18. [26]

    Machine learning: a probabilistic perspective

    Kevin P Murphy. Machine learning: a probabilistic perspective. MIT press, 2012

    2012

  19. [27]

    O'Reilly Media, Inc

    Josh Patterson and Adam Gibson. Deep learning: A practitioner's approach. " O'Reilly Media, Inc.", 2017

    2017

  20. [28]

    Artificial Intelligence: A Modern Approach (4th Edition)

    Stuart Russell and Peter Norvig. Artificial Intelligence: A Modern Approach (4th Edition) . Pearson, 2020. ISBN 9780134610993. URL http://aima.cs.berkeley.edu/

    2020

  21. [29]

    Deep learning

    Ian Goodfellow, Yoshua Bengio, and Aaron Courville. Deep learning. MIT press, 2016

    2016

  22. [30]

    A comprehensive survey of recent trends in deep learning for digital images augmentation

    Nour Eldeen Khalifa, Mohamed Loey, and Seyedali Mirjalili. A comprehensive survey of recent trends in deep learning for digital images augmentation. Artificial Intelligence Review, pages 1--27, 2022

    2022

  23. [31]

    Christopher M. Bishop. Pattern Recognition and Machine Learning. Springer, 2006

    2006

  24. [32]

    Learning deep features for discriminative localization

    Bolei Zhou, Aditya Khosla, Agata Lapedriza, Aude Oliva, and Antonio Torralba. Learning deep features for discriminative localization. In Proceedings of the IEEE conference on computer vision and pattern recognition, pages 2921--2929, 2016

    2016

  25. [33]

    Score-cam: Score-weighted visual explanations for convolutional neural networks

    Haofan Wang, Zifan Wang, Mengnan Du, Fan Yang, Zijian Zhang, Sirui Ding, Piotr Mardziel, and Xia Hu. Score-cam: Score-weighted visual explanations for convolutional neural networks. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition workshops,...

    2020

  26. [34]

    Layercam: Exploring hierarchical class activation maps for localization

    Peng-Tao Jiang, Chang-Bin Zhang, Qibin Hou, Ming-Ming Cheng, and Yunchao Wei. Layercam: Exploring hierarchical class activation maps for localization. IEEE Transactions on Image Processing, 30: 0 5875--5888, 2021

    2021

  27. [35]

    Restricting the flow: Information bottlenecks for attribution

    Karl Schulz, Leon Sixt, Federico Tombari, and Tim Landgraf. Restricting the flow: Information bottlenecks for attribution. arXiv preprint arXiv:2001.00396, 2020

    2001 arXiv

  28. [36]

    Patchnet: interpretable neural networks for image classification

    Adityanarayanan Radhakrishnan, Charles Durham, Ali Soylemezoglu, and Caroline Uhler. Patchnet: interpretable neural networks for image classification. arXiv preprint arXiv:1705.08078, 2017

    2017 arXiv

  29. [37]

    Image data augmentation for deep learning: A survey

    Suorong Yang, Weikang Xiao, Mengcheng Zhang, Suhan Guo, Jian Zhao, and Furao Shen. Image data augmentation for deep learning: A survey. arXiv preprint arXiv:2204.08610, 2022

    2022 arXiv

  30. [38]

    Data augmentation by pairing samples for images classification

    Hiroshi Inoue. Data augmentation by pairing samples for images classification. arXiv preprint arXiv:1801.02929, 2018

    2018 arXiv

  31. [39]

    Autoaugment: Learning augmentation strategies from data

    Ekin D Cubuk, Barret Zoph, Dandelion Mane, Vijay Vasudevan, and Quoc V Le. Autoaugment: Learning augmentation strategies from data. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 113--123, 2019

    2019

  32. [40]

    Learning to compose domain-specific transformations for data augmentation

    Alexander J Ratner, Henry Ehrenberg, Zeshan Hussain, Jared Dunnmon, and Christopher R \'e . Learning to compose domain-specific transformations for data augmentation. Advances in neural information processing systems, 30, 2017

    2017

  33. [41]

    Very deep convolutional neural network based image classification using small training sample size

    Shuying Liu and Weihong Deng. Very deep convolutional neural network based image classification using small training sample size. In 2015 3rd IAPR Asian conference on pattern recognition (ACPR), pages 730--734. IEEE, 2015

    2015

  34. [42]

    A study on cnn transfer learning for image classification

    Mahbub Hussain, Jordan J Bird, and Diego R Faria. A study on cnn transfer learning for image classification. In Advances in Computational Intelligence Systems: Contributions Presented at the 18th UK Workshop on Computational Intelligence, September 5-7, 2018, Nottingham, UK, p...

    2018

  35. [43]

    Deep residual learning for image recognition

    Kaiming He, Xiangyu Zhang, Shaoqing Ren, and Jian Sun. Deep residual learning for image recognition. In Proceedings of the IEEE conference on computer vision and pattern recognition, pages 770--778, 2016

    2016

  36. [44]

    Densely connected convolutional networks

    Gao Huang, Zhuang Liu, Laurens Van Der Maaten, and Kilian Q Weinberger. Densely connected convolutional networks. In Proceedings of the IEEE conference on computer vision and pattern recognition, pages 4700--4708, 2017

    2017

  37. [45]

    What do different evaluation metrics tell us about saliency models? IEEE transactions on pattern analysis and machine intelligence, 41 0 (3): 0 740--757, 2018

    Zoya Bylinskii, Tilke Judd, Aude Oliva, Antonio Torralba, and Fr \'e do Durand. What do different evaluation metrics tell us about saliency models? IEEE transactions on pattern analysis and machine intelligence, 41 0 (3): 0 740--757, 2018

    2018

  38. [46]

    Generalized intersection over union: A metric and a loss for bounding box regression

    Hamid Rezatofighi, Nathan Tsoi, JunYoung Gwak, Amir Sadeghian, Ian Reid, and Silvio Savarese. Generalized intersection over union: A metric and a loss for bounding box regression. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 658--...

    2019

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.