SGD-Mix: Enhancing Domain-Specific Image Classification with Label-Preserving Data Augmentation

Fang-Yi Su; Jung-Hsien Chiang; Yixuan Dong

REVIEW 3 major objections 5 minor 71 references

SGD-Mix: Enhancing Domain-Specific Image Classification with Label-Preserving Data Augmentation

T0 review · 3 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 SGD-Mix is a data augmentation pipeline that keeps the source image's foreground, swaps in a target image's background, and refines the composite with a fine-tuned diffusion model while preserving the source label, reporting top average…

desk verdict Consistent but tiny gains over Diff-Mix, with the label-preserving guarantee asserted rather than measured; a solid candidate for revision, not a breakthrough. read the letter →

arxiv 2505.11813 v1 pith:QM6U3JRI submitted 2025-05-17 cs.CV cs.AI

Yixuan Dong , Fang-Yi Su , Jung-Hsien Chiang This is my paper · ORCID

classification cs.CVcs.AI

keywords dataaugmentationdiffusionmodelssaliency-guidedmixinglabelpreservationfine-grainedclassificationlong-tailfew-shotlearningbackgroundrobustness

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

SGD-Mix is a data augmentation method for domain-specific image classification. It proposes that good augmentation needs diversity, faithfulness, and label clarity at once, and it is designed to supply all three. A saliency map keeps the source image's foreground, a target image supplies the new background, and a fine-tuned diffusion model refines the composite. The paper argues that prior methods such as Diff-Mix and DiffuseMix fail on at least one of these axes, and it reports experiments on fine-grained, long-tail, few-shot, and background-robustness benchmarks where SGD-Mix matches or exceeds the best prior augmentation methods on average accuracy. If the reported gains hold, the method offers a label-preserving alternative to inter-class mixing that avoids semantic drift without external datasets or segmentation models.

What carries the argument

The key machinery is a three-stage pipeline. Saliency-based target selection chooses a target image whose saliency map is closest in $L^2$ distance to the source's, so the background replacement is coherent. Otsu-thresholded binary masks are unioned and the mixed image is formed by pixel-wise composition $I_{(i,j)} = M_{(i,j)} \odot I_i + (1 - M_{(i,j)}) \odot I_j$, keeping the source foreground and replacing the background. A domain-specific diffusion model, fine-tuned with DreamBooth LoRA plus Textual Inversion identifiers of the form '$[v_i]$ $[metaclass]$', refines the composite under the prompt 'a photo of a $[v_s]$ $[metaclass]$', with translation strength $S$ controlling the faithfulness-diversity trade-off. The load-bearing identity is Eq. (11), the asserted label-preserving chain that assigns the source label to every generated image.

What would settle it

Feed the SGD-Mix generated images at the paper's translation strengths ($S = 0.5, 0.7, 0.9$) to a strong pre-trained classifier or human annotators, and compare their labels with the source labels; if a nontrivial fraction land on a different class, or if retraining with verified labels fails to reproduce the reported gains, the label-preserving guarantee is broken and label clarity is not what drives the improvement.

Watch

Extended reading notes

Core claim

The central claim is that composing a source image's saliency-masked foreground with a target image's background, then denoising the composite with a diffusion model fine-tuned on the training domain under a class-specific prompt, yields augmented training images that are diverse, faithful, and unambiguous in label. The paper formalizes label preservation as the chain $\mathrm{Label}(\hat{I}_{(i,j)}) = \mathrm{Label}(I_{(i,j)}) = \mathrm{Label}(I_i)$, asserting that because the mixed image inherits the source label and the diffusion model conditions only on the source class token, the final generated image keeps the source label. Across four task families, the paper reports that this pipeline outperforms prior augmentation methods, including Diff-Mix, DiffuseMix, Mixup, CutMix, and GuidedMixup, with the largest margins on fine-grained dogs and aircraft and on rare classes in long-tail settings.

Load-bearing premise

The load-bearing premise is that the diffusion model's refinement preserves the source foreground's discriminative features closely enough that the assigned source label stays correct, a condition asserted by construction in Eq. (11) and never directly measured.

Editorial extensions

If this is right

Because every generated image carries the source label by construction, downstream classifiers can be trained with standard one-hot labels and label smoothing, without the filtering or label re-estimation that inter-class mixing methods require.
Replacing the background rather than blending it reduces the model's reliance on background cues, which the Waterbird experiment supports with a 6.5-point gain over the baseline on the (waterbird, land) split.
The structured prompt '$[v_i]$ $[metaclass]$' with textual inversion separates fine-grained class identity from the super-category, which should improve generation specificity on datasets with high inter-class similarity.
Because the same pipeline works across fine-grained, long-tail, few-shot, and background-shift settings, the framework is a general recipe for domain-specific augmentation rather than a fix for a single task.

Reading between the lines

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

The paper's spectral-residual ablation suggests saliency choice is not load-bearing, so a cheap saliency estimator might capture most of the gain; the paper does not test this at scale.
If the label-preserving chain bends at high translation strengths, an untested safety valve would be a lightweight classifier-based label check on generated images, re-labeling or discarding those that drift.
SGD-Mix needs only the training dataset itself, so it is naturally portable to domains like medical or satellite imagery where paired external background data do not exist; the paper does not evaluate those settings.
Because background diversity is bounded by the training set, coupling SGD-Mix with an external background pool would probe whether the gains scale; the paper deliberately avoids external datasets.

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.

Desk Editor's Note

read the letter

Here's my read on the SGD-Mix paper. It's a sensible, well-written augmentation method that consistently, but narrowly, improves on Diff-Mix across four task families. The pipeline—saliency-based target selection, union-mask mixing, and diffusion refinement with a class-specific prompt—is a legitimate new combination of known parts. It is not a breakthrough, but it is not a toy either.

The paper's central claim is label clarity: Eq. (11) states Label(Î) = Label(I_mix) = Label(source) by construction. That is the soft spot. The paper itself shows that DiffuseMix suffers semantic drift at high translation strengths (Figure 2), but never measures whether its own pipeline drifts at S=0.7 and 0.9, which are used in experiments. The diffusion model is fine-tuned on all classes, so it has strong class priors; noisy mixed images at high S can plausibly trigger those priors. The attention maps in Figure 4 are illustrative, not quantitative. A simple label-agreement check with a pretrained classifier, or even human evaluation, would settle this. Until that exists, the label-preserving advantage over Diff-Mix is an assumption, not a measured property.

The experimental evidence is otherwise decent in breadth but thin in strength. Most gains over Diff-Mix are between 0.18 and 0.6 points, with no standard deviations or significance tests. Some baseline numbers are borrowed from the Diff-Mix paper; the footnote is honest, but a few of those baselines could be unfair if the settings differed. The ablations on saliency method and batch size are useful, and the limitation note about computational cost is straightforward. No code or data are released, so reproduction requires careful reimplementation.

Overall, this paper deserves a serious referee. The method is plausible, the potential failure mode is informative, and the domain is practically important. For acceptance, the authors should add error bars or significance tests, reproduce the imported baselines, and provide a quantitative label-consistency check across S values. If those hold, it's a fine contribution; if not, the advantage over Diff-Mix may largely evaporate. I'd bring it to reading group only if someone cares about diffusion augmentation, but I'd send it to review.

Referee Report

3 major / 5 minor

Summary. The manuscript introduces SGD-Mix, a three-stage data augmentation framework for domain-specific image classification. In Stage 1, for each source image a target image is selected from a random batch by minimizing the L2 distance between saliency maps; in Stage 2, Otsu-thresholded binary masks are unioned and used to compose an image that keeps the source foreground and replaces the background with the target's; in Stage 3, a DreamBooth/LoRA/Textual-Inversion fine-tuned diffusion model refines the mixed image under a source-class prompt. The authors claim the pipeline guarantees label clarity (Eq. 11) and report experiments on fine-grained, long-tail, few-shot, and background-robustness benchmarks, where they report the highest average accuracy in most settings.

Significance. The contribution is potentially useful: it gives a concrete, well-specified method that targets three acknowledged quality dimensions, and the experimental scope is broad. The saliency-guided target selection and mask-union construction are simple and reproducible from Algorithm 1, and the discussion of why prior methods address only subsets of diversity, faithfulness, and label clarity is clear. However, the central claim of label clarity rests on an asserted rather than measured equality, and the reported gains over Diff-Mix are small. Without uncertainty quantification or a label-consistency check, the strong claims in the abstract are not supported. The paper does not provide code or released data, so reproducibility currently depends on the written protocol alone.

major comments (3)

[§5.3, Eq. (11)] The equality Label(Î(i,j)) = Label(I(i,j)) = Label(I_i) is asserted by construction, but the paper never empirically verifies label consistency of the refined images. This is load-bearing because label clarity is the paper's distinguishing advantage over Diff-Mix's nonlinear label mixing and over DiffuseMix's drift-prone refinement. The authors themselves document in Fig. 2 that diffusion-based transformation can cause semantic drift at high translation strengths, so the same risk must be quantified for SGD-Mix, especially at S = 0.7 and S = 0.9 used in the experiments. I request a label-consistency evaluation, for example agreement between source labels and predictions of a strong classifier, or human evaluation, reported separately for each task family; the union mask construction alone does not rule out residual target-foreground artifacts or diffusion-induced semantic changes.
[Tables 2–5] The reported superiority over Diff-Mix is very small: Table 2 shows average gains of +0.18% (ResNet50) and +0.29% (ViT), Table 5 shows +0.02% on Waterbird, and long-tail gains are around 0.5–1.5 points on some IF settings. No standard deviations, number of runs, confidence intervals, or significance tests are reported. Given these small deltas, the abstract's claim of superior performance over state-of-the-art approaches is not statistically supported. Please provide repeated-run statistics and significance tests, and identify explicitly which per-dataset or per-settings comparisons are robust and which are within noise.
[§6.1–§6.3] The method has several free hyperparameters (N, expansion multiplier, replacement probability p, label-smoothing confidence, and translation strength S), and S is set differently across experiments: 0.5/0.7/0.9 for fine-grained, 0.7 for long-tail, and 0.9 for few-shot. The main text ablates only N (Fig. 6) and the saliency method (Table 6); there is no accuracy-based sensitivity study for S, which is the parameter that directly controls the diversity-faithfulness trade-off and whose choice is part of the method's claimed flexibility. Please report accuracy over a sweep of S for at least one dataset in each task family, or provide a principled criterion for the per-task choice.

minor comments (5)

[§5.3] The prompt is written as "a photo of a [vs] [metaclass]" but the text later says "where vs is the learnable embedding"; please use consistent bracket notation, e.g., [vs], to avoid confusion.
[Table 2] The baseline row is formatted with a leading dash and the numbers appear shifted relative to the method column; please align rows so each column corresponds to one dataset.
[§6.1, footnote 2] The phrase "others are reported by Diff-Mix [64], slightly outperforming ours under consistent settings" is confusing; please clarify whether the cited numbers are higher or lower than the authors' reproductions and exactly which hyperparameters and protocols were matched.
[§6.1–§6.4] The method names "Da-Fusion" and "DA-fusion" are used inconsistently across the text and Tables 3–5; please standardize.
[Figure 5] The visual S sweep is informative, but a quantitative measure of diversity or faithfulness (e.g., LPIPS or FID of the generated sets, or classifier agreement with source labels) would better support the claimed balance between diversity and faithfulness.

Circularity Check

0 steps flagged · score 0.0 of 10

No circular derivation: Eq. (11) is a label-assignment rule, and the central empirical claims are tested on external benchmarks.

full rationale

The derivation chain is self-contained. SGD-Mix's construction (Eqs. 7-10) deterministically produces a mixed image whose foreground is the source image's saliency region, and Eq. 11 then assigns the source label to the refined image. This is a label-assignment rule, not a parameter fitted to data, and not a prediction derived from a model whose inputs include the target quantity. The claim of label clarity is definitional in the sense that the method defines its own labeling, but the paper's central empirical claim—that classifiers trained with SGD-Mix augmentations outperform baselines—is tested on held-out external benchmarks (CUB, Cars, Flowers, Dogs, Aircraft, Waterbird) against published and reproduced competitor numbers. No load-bearing self-citation appears: references to Diff-Mix and DiffuseMix are external prior work, and the paper does not invoke a uniqueness theorem or imported ansatz to rule out alternatives. The only weakness is that Eq. 11's label-preservation is not directly empirically verified (e.g., no label-consistency accuracy on generated images), which is a correctness/validation risk, not a circularity. Hence score 0.

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

The method depends on standard diffusion-model fine-tuning and saliency-map techniques from prior work; the main free parameters are augmentation hyperparameters tuned per dataset. No new physical or model entities are introduced.

free parameters (5)

Translation strength S = 0.5/0.7/0.9 for fine-grained, 0.7 for long-tail, 0.9 for few-shot
Controls noise injection in diffusion refinement; tuned per task and reported in Sections 6.1, 6.2, and 6.3.
Target batch size N = 50
Number of candidate targets sampled for saliency matching; ablation in Figure 6 shows performance saturating near 50.
Expansion multiplier = 5
Number of augmented samples generated per real image; used in Sections 6.1 and 6.3.
Replacement probability p = 0.1 for fine-grained/long-tail/all-shot; 0.5, 0.3, 0.2 for 1, 5, 10-shot
Probability of replacing a training image with its augmented counterpart; varies by experiment.
Label smoothing confidence = 0.9
Used in fine-grained training (Section 6.1) to soften targets.

assumptions (3)

domain assumption Saliency masks from Grad-CAM and Otsu thresholding correctly separate foreground from background.
Used in Section 5.2 to build the mixing mask; if masks are inaccurate, the mixed image may contain foreign foreground or lose the source foreground.
domain assumption The fine-tuned diffusion model preserves source-class semantics under image-to-image translation at the chosen S.
Invoked in Section 5.3 and Eq. (11) for label preservation; no quantitative verification of generated-image label consistency is provided.
domain assumption The benchmark datasets and evaluation protocols are representative and correctly followed.
The paper relies on standard protocols for CUB, Cars, Flowers, Dogs, Aircraft, and Waterbird; deviations would affect the reported accuracy.

how reviews work

0 comments

Cite this review

Pith. "Pith review of SGD-Mix: Enhancing Domain-Specific Image Classification with Label-Preserving Data Augmentation." pith.science (2026). https://pith.science/paper/QM6U3JRI

@misc{pith2026250511813,
  author       = {Pith},
  title        = {Pith review of: SGD-Mix: Enhancing Domain-Specific Image Classification with Label-Preserving Data Augmentation},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/QM6U3JRI}},
  note         = {Machine review of arXiv:2505.11813}
}

read the original abstract

Data augmentation for domain-specific image classification tasks often struggles to simultaneously address diversity, faithfulness, and label clarity of generated data, leading to suboptimal performance in downstream tasks. While existing generative diffusion model-based methods aim to enhance augmentation, they fail to cohesively tackle these three critical aspects and often overlook intrinsic challenges of diffusion models, such as sensitivity to model characteristics and stochasticity under strong transformations. In this paper, we propose a novel framework that explicitly integrates diversity, faithfulness, and label clarity into the augmentation process. Our approach employs saliency-guided mixing and a fine-tuned diffusion model to preserve foreground semantics, enrich background diversity, and ensure label consistency, while mitigating diffusion model limitations. Extensive experiments across fine-grained, long-tail, few-shot, and background robustness tasks demonstrate our method's superior performance over state-of-the-art approaches.

Figures

Figures reproduced from arXiv: 2505.11813 by the authors.

**Figure 2.** Semantic drift in DiffuseMix [20] generated images using the prompt “A transformed version of image into autumn”. Stronger transformations reduce semantic fidelity to the source image. More examples in Supplementary Materials 2. where θ \ ϕ denotes the frozen parameters and ϕ the lowrank adapted subset. By restricting the model updates to the LoRA components within the U-Net, we retain much of the original diffusio… view at source ↗

**Figure 3.** The pipeline of the proposed SGD-Mix method. The process involves three major stages: (1) saliency-based target selection, (2) [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗

Figures from the paper (2 more)

**Figure 4.** Figure 4: Visualization of attention maps before and after saliency [PITH_FULL_IMAGE:figures/full_fig_p005_4.png]

**Figure 6.** Figure 6: Accuracy on CUB with ResNet50, varying N (Section 6.1). SGD-Mix improves with N, saturating near 50. match foregrounds and backgrounds within a dataset without an optimal pairing mechanism. In contrast, SGD-Mix uses L2 distance between saliency maps to pair similar f…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

71 extracted references · 61 canonical work pages

[1]

Antoniou, A

A. Antoniou, A. Storkey, and H. Edwards. Augmenting image classifiers using data augmentation generative adversarial networks. InInternational conference on artificial neural networks, pages 594–603. Springer, 2018

work page 2018
[2]

Antoniou, A

A. Antoniou, A. Storkey, and H. Edwards. Data augmentation generative adversarial networks.stat, 1050:8, 2018

work page 2018
[3]

Araslanov and S

N. Araslanov and S. Roth. Self-supervised augmentation con- sistency for adapting semantic segmentation. InProceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 15384–15394, 2021

work page 2021
[4]

Azizi, S

S. Azizi, S. Kornblith, C. Saharia, M. Norouzi, and D. J. Fleet. Synthetic data from diffusion models improves imagenet clas- sification.Transactions on Machine Learning Research

work page
[5]

Brooks, A

T. Brooks, A. Holynski, and A. A. Efros. Instructpix2pix: Learning to follow image editing instructions. InProceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 18392–18402, 2023

work page 2023
[6]

K. Cao, C. Wei, A. Gaidon, N. Arechiga, and T. Ma. Learning imbalanced datasets with label-distribution-aware margin loss. Advances in neural information processing systems, 32, 2019

work page 2019
[7]

Chapelle, J

O. Chapelle, J. Weston, L. Bottou, and V . Vapnik. Vicinal risk minimization.Advances in neural information processing systems, 13, 2000

work page 2000
[8]

T. Chen. On the importance of noise scheduling for diffusion models.arXiv preprint arXiv:2301.10972, 2023

arXiv 2023

Show all 71 references

[9]

Dhariwal and A

P. Dhariwal and A. Nichol. Diffusion models beat gans on image synthesis.Advances in neural information processing systems, 34:8780–8794, 2021

2021
[10]

Dosovitskiy, L

A. Dosovitskiy, L. Beyer, A. Kolesnikov, D. Weissenborn, X. Zhai, T. Unterthiner, M. Dehghani, M. Minderer, G. Heigold, S. Gelly, et al. An image is worth 16x16 words: Transformers for image recognition at scale. InInternational Conference on Learning Representations, 2020

2020
[11]

R. Gal, Y . Alaluf, Y . Atzmon, O. Patashnik, A. H. Bermano, G. Chechik, and D. Cohen-or. An image is worth one word: Personalizing text-to-image generation using textual inver- sion. InThe Eleventh International Conference on Learning Representations
[12]

Ghiasi, Y

G. Ghiasi, Y . Cui, A. Srinivas, R. Qian, T.-Y . Lin, E. D. Cubuk, Q. V . Le, and B. Zoph. Simple copy-paste is a strong data augmentation method for instance segmentation. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 2918–2928, 2021

2021
[13]

Goodfellow, J

I. Goodfellow, J. Pouget-Abadie, M. Mirza, B. Xu, D. Warde- Farley, S. Ozair, A. Courville, and Y . Bengio. Generative adversarial nets.Advances in neural information processing systems, 27, 2014

2014
[14]

K. He, X. Zhang, S. Ren, and J. Sun. Deep residual learning for image recognition. InProceedings of the IEEE conference on computer vision and pattern recognition, pages 770–778, 2016

2016
[15]

R. He, S. Sun, X. Yu, C. Xue, W. Zhang, P. Torr, S. Bai, and X. QI. Is synthetic data from generative models ready for image recognition? InThe Eleventh International Conference on Learning Representations
[16]

J. Ho, A. Jain, and P. Abbeel. Denoising diffusion proba- bilistic models.Advances in neural information processing systems, 33:6840–6851, 2020

2020
[17]

Hou and L

X. Hou and L. Zhang. Saliency detection: A spectral residual approach. In2007 IEEE Conference on computer vision and pattern recognition, pages 1–8. Ieee, 2007

2007
[18]

E. J. Hu, Y . Shen, P. Wallis, Z. Allen-Zhu, Y . Li, S. Wang, L. Wang, W. Chen, et al. Lora: Low-rank adaptation of large language models.ICLR, 1(2):3, 2022

2022
[19]

Huang, X

S. Huang, X. Wang, and D. Tao. Snapmix: Semantically proportional mixing for augmenting fine-grained data. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 35, pages 1628–1636, 2021

2021
[20]

Islam, M

K. Islam, M. Z. Zaheer, A. Mahmood, and K. Nandakumar. Diffusemix: Label-preserving data augmentation with diffu- sion models. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 27621– 27630, 2024

2024
[21]

L. Itti, C. Koch, and E. Niebur. A model of saliency-based visual attention for rapid scene analysis.IEEE Transactions on pattern analysis and machine intelligence, 20(11):1254– 1259, 2002

2002
[22]

Kang and S

M. Kang and S. Kim. Guidedmixup: an efficient mixup strategy guided by saliency maps. InProceedings of the AAAI conference on artificial intelligence, volume 37, pages 1096–1104, 2023

2023
[23]

Khosla, N

A. Khosla, N. Jayadevaprakash, B. Yao, and F.-F. Li. Novel dataset for fine-grained image categorization: Stanford dogs. InProc. CVPR workshop on fine-grained visual categoriza- tion (FGVC), volume 2, 2011

2011
[24]

J.-H. Kim, W. Choo, and H. O. Song. Puzzle mix: Exploiting saliency and local statistics for optimal mixup. InInterna- tional conference on machine learning, pages 5275–5285. PMLR, 2020

2020
[25]

D. P. Kingma, M. Welling, et al. Auto-encoding variational bayes
[26]

Krause, M

J. Krause, M. Stark, J. Deng, and L. Fei-Fei. 3d object repre- sentations for fine-grained categorization. InProceedings of the IEEE international conference on computer vision work- shops, pages 554–561, 2013

2013
[27]

Kumari, B

N. Kumari, B. Zhang, R. Zhang, E. Shechtman, and J.-Y . Zhu. Multi-concept customization of text-to-image diffusion. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 1931–1941, 2023

1931
[28]

Litjens, T

G. Litjens, T. Kooi, B. E. Bejnordi, A. A. A. Setio, F. Ciompi, M. Ghafoorian, J. A. Van Der Laak, B. Van Ginneken, and C. I. S´anchez. A survey on deep learning in medical image analysis.Medical image analysis, 42:60–88, 2017. 9

2017
[29]

Z. Liu, Z. Miao, X. Zhan, J. Wang, B. Gong, and S. X. Yu. Large-scale long-tailed recognition in an open world. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 2537–2546, 2019

2019
[30]

S. Maji, E. Rahtu, J. Kannala, M. Blaschko, and A. Vedaldi. Fine-grained visual classification of aircraft. 2013

2013
[31]

C. Meng, Y . He, Y . Song, J. Song, J. Wu, J.-Y . Zhu, and S. Ermon. Sdedit: Guided image synthesis and editing with stochastic differential equations. InInternational Conference on Learning Representations
[32]

M¨uller, S

R. M¨uller, S. Kornblith, and G. E. Hinton. When does label smoothing help?Advances in neural information processing systems, 32, 2019

2019
[33]

Mumuni and F

A. Mumuni and F. Mumuni. Data augmentation: A com- prehensive survey of modern approaches.Array, 16:100258, 2022

2022
[34]

A. Q. Nichol and P. Dhariwal. Improved denoising diffusion probabilistic models. InInternational conference on machine learning, pages 8162–8171. PMLR, 2021

2021
[35]

A. Q. Nichol, P. Dhariwal, A. Ramesh, P. Shyam, P. Mishkin, B. Mcgrew, I. Sutskever, and M. Chen. Glide: Towards photorealistic image generation and editing with text-guided diffusion models. InInternational Conference on Machine Learning, pages 16784–16804. PMLR, 2022

2022
[36]

Nilsback and A

M.-E. Nilsback and A. Zisserman. Automated flower classifi- cation over a large number of classes. In2008 Sixth Indian conference on computer vision, graphics & image processing, pages 722–729. IEEE, 2008

2008
[37]

Otsu et al

N. Otsu et al. A threshold selection method from gray-level histograms.Automatica, 11(285-296):23–27, 1975

1975
[38]

S. Park, Y . Hong, B. Heo, S. Yun, and J. Y . Choi. The majority can help the minority: Context-rich minority oversampling for long-tailed classification. InProceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 6887–6896, 2022

2022
[39]

Patel, T

M. Patel, T. Gokhale, C. Baral, and Y . Yang. Conceptbed: Evaluating concept learning abilities of text-to-image diffu- sion models. InProceedings of the AAAI Conference on Artificial Intelligence, volume 38, pages 14554–14562, 2024

2024
[40]

T. Qi, S. Fang, Y . Wu, H. Xie, J. Liu, L. Chen, Q. He, and Y . Zhang. Deadiff: An efficient stylization diffusion model with disentangled representations. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 8693–8702, 2024

2024
[41]

H. Qin, X. Jin, H. Zhu, H. Liao, M. A. El-Yacoubi, and X. Gao. Sumix: Mixup with semantic and uncertain infor- mation. InEuropean Conference on Computer Vision, pages 70–88. Springer, 2024

2024
[42]

J. Qin, J. Fang, Q. Zhang, W. Liu, X. Wang, and X. Wang. Resizemix: Mixing data with preserved object information and true labels.arXiv preprint arXiv:2012.11101, 2020

2012 arXiv
[43]

Z. Qiu, W. Liu, H. Feng, Y . Xue, Y . Feng, Z. Liu, D. Zhang, A. Weller, and B. Sch¨olkopf. Controlling text-to-image diffu- sion by orthogonal finetuning.Advances in Neural Informa- tion Processing Systems, 36:79320–79362, 2023

2023
[44]

Radford, J

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. InInternational conference on machine learning, pages 8748–8763. PmLR, 2021

2021
[45]

Ridnik, E

T. Ridnik, E. Ben-Baruch, A. Noy, and L. Zelnik-Manor. Imagenet-21k pretraining for the masses. InThirty-fifth Con- ference on Neural Information Processing Systems Datasets and Benchmarks Track (Round 1)
[46]

Rombach, A

R. Rombach, A. Blattmann, D. Lorenz, P. Esser, and B. Om- mer. High-resolution image synthesis with latent diffusion models. InProceedings of the IEEE/CVF conference on com- puter vision and pattern recognition, pages 10684–10695, 2022

2022
[47]

Ronneberger, P

O. Ronneberger, P. Fischer, and T. Brox. U-net: Convolutional networks for biomedical image segmentation. InMedical im- age computing and computer-assisted intervention–MICCAI 2015: 18th international conference, Munich, Germany, Oc- tober 5-9, 2015, proceedings, part III 18, ...

2015
[48]

Rother, V

C. Rother, V . Kolmogorov, and A. Blake. ” grabcut” inter- active foreground extraction using iterated graph cuts.ACM transactions on graphics (TOG), 23(3):309–314, 2004

2004
[49]

N. Ruiz, Y . Li, V . Jampani, Y . Pritch, M. Rubinstein, and K. Aberman. Dreambooth: Fine tuning text-to-image diffu- sion models for subject-driven generation. InProceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 22500–22510, 2023

2023
[50]

Russakovsky, J

O. Russakovsky, J. Deng, H. Su, J. Krause, S. Satheesh, S. Ma, Z. Huang, A. Karpathy, A. Khosla, M. Bernstein, et al. Ima- genet large scale visual recognition challenge.International journal of computer vision, 115:211–252, 2015

2015
[51]

Sagawa, P

S. Sagawa, P. W. Koh, T. B. Hashimoto, and P. Liang. Distribu- tionally robust neural networks. InInternational Conference on Learning Representations
[52]

Saharia, W

C. Saharia, W. Chan, S. Saxena, L. Li, J. Whang, E. L. Den- ton, K. Ghasemipour, R. Gontijo Lopes, B. Karagol Ayan, T. Salimans, et al. Photorealistic text-to-image diffusion mod- els with deep language understanding.Advances in neural information processing systems, 35:36479–...

2022
[53]

Samuel, Y

D. Samuel, Y . Atzmon, and G. Chechik. From generalized zero-shot learning to long-tail with class descriptors. InPro- ceedings of the IEEE/CVF winter conference on applications of computer vision, pages 286–295, 2021

2021
[54]

R. R. Selvaraju, M. Cogswell, A. Das, R. Vedantam, D. Parikh, and D. Batra. Grad-cam: Visual explanations from deep networks via gradient-based localization. InPro- ceedings of the IEEE international conference on computer vision, pages 618–626, 2017

2017
[55]

Sengupta, V

S. Sengupta, V . Jayaram, B. Curless, S. M. Seitz, and I. Kemelmacher-Shlizerman. Background matting: The world is your green screen. InProceedings of the IEEE/CVF Con- ference on Computer Vision and Pattern Recognition, pages 2291–2300, 2020

2020
[56]

Shorten and T

C. Shorten and T. M. Khoshgoftaar. A survey on image data augmentation for deep learning.Journal of big data, 6(1):1– 48, 2019

2019
[57]

Simonyan, A

K. Simonyan, A. Vedaldi, and A. Zisserman. Deep inside convolutional networks: visualising image classification mod- 10 els and saliency maps. International Conference on Learning Representations, 2014

2014
[58]

J. Song, C. Meng, and S. Ermon. Denoising diffusion implicit models. InInternational Conference on Learning Represen- tations
[59]

Y . Song, J. Sohl-Dickstein, D. P. Kingma, A. Kumar, S. Er- mon, and B. Poole. Score-based generative modeling through stochastic differential equations. InInternational Conference on Learning Representations
[60]

Trabucco, K

B. Trabucco, K. Doherty, M. Gurinas, and R. Salakhutdinov. Effective data augmentation with diffusion models. InR0- FoMo: Robustness of Few-shot and Zero-shot Learning in Large Foundation Models
[61]

Tumanyan, M

N. Tumanyan, M. Geyer, S. Bagon, and T. Dekel. Plug- and-play diffusion features for text-driven image-to-image translation. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 1921–1930, 2023

1921
[62]

A. S. Uddin, M. S. Monira, W. Shin, T. Chung, and S.-H. Bae. Saliencymix: A saliency guided data augmentation strategy for better regularization. InInternational Conference on Learning Representations
[63]

C. Wah, S. Branson, P. Welinder, P. Perona, and S. Belongie. The caltech-ucsd birds-200-2011 dataset. 2011

2011
[64]

Z. Wang, L. Wei, T. Wang, H. Chen, Y . Hao, X. Wang, X. He, and Q. Tian. Enhance image classification via inter- class image mixup with diffusion model. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 17223–17233, 2024

2024
[65]

Ye-Bin, N

M. Ye-Bin, N. Hyeon-Woo, W. Choi, N. Kim, S. Kwak, and T.-H. Oh. Exploiting synthetic data for data imbalance problems: baselines from a data perspective.arXiv preprint arXiv:2308.00994, 6, 2023

2023 arXiv
[66]

J. Yu, Z. Lin, J. Yang, X. Shen, X. Lu, and T. S. Huang. Generative image inpainting with contextual attention. In Proceedings of the IEEE conference on computer vision and pattern recognition, pages 5505–5514, 2018

2018
[67]

S. Yun, D. Han, S. J. Oh, S. Chun, J. Choe, and Y . Yoo. Cutmix: Regularization strategy to train strong classifiers with localizable features. InProceedings of the IEEE/CVF international conference on computer vision, pages 6023– 6032, 2019

2019
[68]

Zhang, M

H. Zhang, M. Cisse, Y . N. Dauphin, and D. Lopez-Paz. mixup: Beyond empirical risk minimization. InInternational Confer- ence on Learning Representations, 2018

2018
[69]

R. Zhao, W. Ouyang, H. Li, and X. Wang. Saliency detection by multi-context deep learning. InProceedings of the IEEE conference on computer vision and pattern recognition, pages 1265–1274, 2015

2015
[70]

B. Zhou, A. Khosla, A. Lapedriza, A. Oliva, and A. Torralba. Learning deep features for discriminative localization. In Proceedings of the IEEE conference on computer vision and pattern recognition, pages 2921–2929, 2016

2016
[71]

B. Zhou, A. Lapedriza, A. Khosla, A. Oliva, and A. Torralba. Places: A 10 million image database for scene recognition. IEEE transactions on pattern analysis and machine intelli- gence, 40(6):1452–1464, 2017. 11

2017

Pith tools

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

[1] [1]

Antoniou, A

A. Antoniou, A. Storkey, and H. Edwards. Augmenting image classifiers using data augmentation generative adversarial networks. InInternational conference on artificial neural networks, pages 594–603. Springer, 2018

work page 2018

[2] [2]

Antoniou, A

A. Antoniou, A. Storkey, and H. Edwards. Data augmentation generative adversarial networks.stat, 1050:8, 2018

work page 2018

[3] [3]

Araslanov and S

N. Araslanov and S. Roth. Self-supervised augmentation con- sistency for adapting semantic segmentation. InProceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 15384–15394, 2021

work page 2021

[4] [4]

Azizi, S

S. Azizi, S. Kornblith, C. Saharia, M. Norouzi, and D. J. Fleet. Synthetic data from diffusion models improves imagenet clas- sification.Transactions on Machine Learning Research

work page

[5] [5]

Brooks, A

T. Brooks, A. Holynski, and A. A. Efros. Instructpix2pix: Learning to follow image editing instructions. InProceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 18392–18402, 2023

work page 2023

[6] [6]

K. Cao, C. Wei, A. Gaidon, N. Arechiga, and T. Ma. Learning imbalanced datasets with label-distribution-aware margin loss. Advances in neural information processing systems, 32, 2019

work page 2019

[7] [7]

Chapelle, J

O. Chapelle, J. Weston, L. Bottou, and V . Vapnik. Vicinal risk minimization.Advances in neural information processing systems, 13, 2000

work page 2000

[8] [8]

T. Chen. On the importance of noise scheduling for diffusion models.arXiv preprint arXiv:2301.10972, 2023

arXiv 2023