Pith. sign in

REVIEW 3 major objections 4 minor 51 references

GIFDL: Generated Image Fluctuation Distortion Learning for Enhancing Steganographic Security

T0 review · 3 major / 4 minor · reviewed 2026-08-16 · deepseek-v4-flash

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

Pith's one-line read The paper claims that the tiny 'fluctuation' differences between near-identical regenerations of a generated image provide a cover distribution that a GAN can learn, letting stego images pass as benign AI-generation noise and raising…

desk verdict Solid incremental method paper whose main number is weakened by test-set tuning and an unreported threshold; worth review but the mechanism is underdetermined. read the letter →

arxiv 2504.15139 v1 pith:TARPMBEY submitted 2025-04-21 cs.CR

Xiangkun Wang , Kejiang Chen , Yuang Qi , Ruiheng Liu , Weiming Zhang , Nenghai Yu This is my paper · ORCID
classification cs.CR
keywords steganographysteganalysisgeneratedimagesdistortionlearninggenerativeadversarialnetworksfluctuationStableDiffusionvolatilitycost
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

GIFDL aims to prove that the pixel-level 'fluctuation' of generated images—the tiny differences between two near-identical regenerations of the same prompt—can be turned into a steganographic distortion model. The paper trains a GAN so that stego images resemble not only their cover but also a randomly chosen fluctuation image, effectively converting embedding modifications into the same benign noise the generative model produces anyway. This matters because it offers a black-box route to more secure steganography in AI-generated images: only the ability to sample slight variations is needed, not access to the generator's internals. In the reported evaluation, the method raises detection-error rates by an average of 3.30% across three steganalyzers compared with the strongest GAN-based baseline, and it retains most of its advantage on real-user Stable Diffusion images while degrading on Midjourney images.

What carries the argument

The central object is the 'fluctuation image': a near-duplicate of the cover produced from the same prompt and seed with a slightly altered CFG scale, which the paper treats as a sample from the generation model's natural pixel noise. GIFDL filters these samples with an MSE threshold τ (Eq. 1), so only fluctuation images sufficiently close to the cover enter training; one such image is then selected at random each epoch and fed to a second discriminator. The key training device is the 'assignment' strategy: Discriminator #1 (Yed-Net) judges cover-vs-stego, Discriminator #2 (XuNet) judges fluctuation-vs-stego, and each iteration updates only the weaker discriminator, preventing the fluctuation-vs-stego task from collapsing to zero loss and starving the generator of gradients. A U-Net generator with a double-tanh embedding simulator converts the learned probabilities into embedding costs, and STC performs the actual payload embedding.

What would settle it

Re-run the training with several different values of τ (including no filtering at all) on the same cover and fluctuation sets, and measure SRM, CovNet, and LWENet error rates at 0.4 bpp. If error rates are flat across very different τ values, the threshold is not the mechanism; if they move strongly with τ, the fluctuation-distribution hypothesis is load-bearing and the missing τ value should be reported for exact reproduction.

Watch

Extended reading notes

Core claim

The central claim is that steganographic distortion can be learned from the fluctuation distribution of black-box generated images rather than imposed by a manually designed cost or by adversarial pressure alone. For each cover image, GIFDL generates several near-identical fluctuation images by slightly changing the CFG scale, then trains a U-Net generator with two steganalysis discriminators: one that separates cover from stego and one that separates a random fluctuation image from stego. Training the generator against both forces the stego image to stay close to the set of images the model would naturally produce, which the paper argues is exactly the right cover-source distribution. The experiments report that this raises the average detection error of SRM, CovNet, and LWENet by 3.30% relative to GMAN, and that combining GIFDL with volatility cost adds a further 9.16% average gain, indicating the learned fluctuation model captures something the Gaussian volatility approximation misses.

Load-bearing premise

The load-bearing premise is that the pixel-wise distribution of fluctuation images—obtained by slightly varying the CFG scale and then filtering with an unreported threshold τ—marks exactly which pixel modifications are safe, so that a stego image close to that distribution becomes hard for steganalyzers to detect; if that distribution is not representative of undetectable modifications, the method's advantage collapses.

Editorial extensions

If this is right

  • No access to the black box is needed: a user of a deployed text-to-image model can generate cover images plus fluctuation samples and train the distortion model using only the model's outputs.
  • The learned distortion transfers, at least partly, across generator variants: training on one Stable Diffusion setup improves detection-error rates on real-user DiffusionDB images, and the advantage persists though weaker on Midjourney-generated images.
  • Adding a hand-modeled volatility cost on top of GIFDL improves average detection error by 9.16% over GIFDL alone, suggesting the learned fluctuation distribution and the Gaussian volatility model are complementary rather than redundant.
  • At higher payloads, giving more weight to the fluctuation-image discriminator can help because the fluctuation distribution offers additional secure modification regions, so the method scales in usefulness as message length grows.

Reading between the lines

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

  • Our inference: the same fluctuation-as-cover-noise mechanism is not limited to CFG scales in diffusion models; any generative model with a continuous sampling knob (sampling temperature, truncation threshold, noise schedule) should exhibit comparable near-duplicate fluctuations, so the recipe could generalize to other generators.
  • Our inference: since GIFDL+VC outperforms GIFDL, a direct learned estimator of the per-pixel volatility distribution—estimated from fluctuation samples without a Gaussian assumption—could plausibly absorb the hand-designed volatility cost into a single network and improve further.
  • Our inference: the paper never reports the numerical value of the threshold τ, so the public comparison cannot be reproduced exactly; sweeping τ and reporting it would let readers separate the contribution of fluctuation-image filtering from the contribution of the two-discriminator adversarial training.
  • Our inference: the weaker transfer to Midjourney images suggests each generator family has its own fluctuation signature, so practical deployment would require sampling fluctuations from the specific target model and re-tuning τ per model.
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

3 major / 4 minor

Summary. The paper proposes GIFDL, a steganographic cost-learning method for images generated by text-to-image diffusion models. A U-Net generator produces modification probabilities from a cover image, and a text-to-image model (Stable Diffusion) is used to create a cover plus ten "fluctuation" images by slightly varying the CFG-scale parameter; a threshold τ filters which fluctuation images are used. Two steganalysis networks are assigned to separate discrimination tasks—cover/stego and fluctuation/stego—and are updated alternately, while the generator is trained against both. Embedding costs are derived from the learned probabilities and messages are embedded with STC. Experiments on a custom INtrain/INtest dataset and on DiffusionDB and JourneyDB report detection error rates for SRM, CovNet, and LWENet, and the paper claims an average 3.30 percentage point improvement over GMAN. The central mechanism is that making stego images statistically close to fluctuation images should improve security against steganalysis.

Significance. If the central claim were fully supported, GIFDL would be a practical and conceptually interesting advance: it uses fluctuation statistics of black-box generated images to shape steganographic distortion, it is evaluated on large public-style datasets, and it is shown to combine with the existing volatility-cost approach. The paper also provides a useful ablation structure separating the roles of fluctuation images, the threshold, and the assigned-discriminator strategy. However, the strength of the empirical evidence is currently weakened by three coupled issues: the main result is measured on the same test set used to select λ, the key threshold τ is never reported or varied, and no run-to-run variability is reported. These issues need to be remedied before the security improvement can be considered established.

major comments (3)
  1. [Section IV-C1, Tables I and II] The main performance claim is supported by a model-selection loop on the same test set. Section IV-C1 tunes λ on INtest by training steganalyzers on 10,000 INtest cover-stego pairs split into 4,000/1,000/5,000 parts, and Table I reports the resulting detection error rates. Section IV-C2 then reports GIFDL in Table II with exactly the same numbers as Table I's λ=1 row (for example, SRM 33.99, CovNet 29.40, LWENet 30.52 at 0.4 bpp) on the same INtest dataset and the same split configuration. Thus the claimed 3.30% average gain over GMAN is measured on the data used to choose λ, so it is not an independent evaluation and may be inflated by selection bias. Please select λ on a separate validation set and re-evaluate on an untouched test set, or present the full λ sweep as the main result with appropriate multiple-testing caveats.
  2. [Eq. (1), Sections III-B1 and IV-C1] The threshold τ is load-bearing but is never specified or varied. Eq. (1) filters fluctuation images using MSE(C,F_k)≤τ, and the ablation contrast between Variant #2 (no threshold) and Variant #3 (threshold) is the main evidence that the fluctuation signal helps. However, the paper never reports the numerical value of τ, and Section IV-C1, despite its title "The optimal value of the parameter τ", actually tunes λ only. Please report τ, describe how many fluctuation images pass the threshold, and include a sensitivity analysis over τ. In addition, to establish that the gain comes from the fluctuation distribution rather than from supplying an auxiliary discriminator with near-duplicate covers, please add a control in which fluctuation images are replaced by generic near-duplicate perturbations of matched MSE.
  3. [Tables I-VII] No measure of run-to-run variability is provided for any reported number. All detection error rates are single values with no standard deviations or repeated-seed runs, and the claimed advantage over the closest competitor is small in several conditions (for example, GACL 33.68 vs GIFDL 33.99 for SRM at 0.4 bpp in Table II, a gap of 0.31 percentage points). Since GAN-based cost learning and steganalyzer training are stochastic, the reported 3.30% average gap cannot be distinguished from training-run variance. Please report means and standard deviations over multiple independent training runs with different seeds, and use a paired or otherwise appropriate statistical comparison where feasible.
minor comments (4)
  1. [Section II-B] The sentence beginning "Since VC calculates distortion only in terms of image volatility, for comprehensive consideration, Since VC calculates distortion solely in terms of image volatility..." is a duplicated and ungrammatical construction; it should be rewritten.
  2. [Eq. (2) and Section III-B1] The symbol N is overloaded: it denotes the number of fluctuation images (set to 10) in Section III-B1, but in Eq. (2) it denotes the total number of pixels. Please use a different symbol, such as H×W, for the pixel count.
  3. [Section I, Contributions] The first contribution item contains a typo: "we take the fluctuation images as as the input of our network"; the duplicated "as" should be removed.
  4. [Throughout] The name of the steganalysis network from reference [26] is written inconsistently as "Xu-Net" and "XuNet"; please standardize the spelling.

Circularity Check

1 steps flagged · score 5.0 of 10

Headline 3.30% security gain is reported on the same INtest set used to select λ, so the headline number is an in-sample optimum; the fluctuation-distortion mechanism itself is not circular.

  1. fitted input called prediction [Section IV-C1 (parameter selection on INtest) feeding Section IV-C2/Table II (headline security comparison)]
    "After training GIFDL, we use 10,000 grayscale images from the INtest dataset as the cover images for testing the security of GIFDL. ... The experimental results, shown in Table I, indicate that when λ = 1 , GIFDL exhibits the best resistance against the three steganalyzers, achieving optimal security performance. Therefore, in all subsequent experiments, we set the parameterλ to 1 for GIFDL."

    λ is selected by evaluating detection error rates on INtest, and the paper's headline 3.30% average improvement over GMAN is then reported from Table II on the same INtest set, with Table II's GIFDL rows identical to Table I's λ=1 rows. Thus the central quantitative claim is the best of the tried λ values on the test set, not an independent prediction on held-out data. This is partial rather than total circularity because the security gain persists for λ=2, 4, and 8 and on external datasets, so the method has content beyond this one selected configuration.

full rationale

GIFDL's core derivation is self-contained: fluctuation images are defined by CFG-perturbed generations, the U-Net generator converts cover images into modification probabilities, and the two discriminators encourage stego images to resemble cover and fluctuation images. No equation in the method section defines the fluctuation target in terms of the final security metric, and the actual embedding uses STC with costs derived from the learned probabilities, so the method is not circular by construction. The main circularity is in the evaluation protocol: Section IV-C1 tunes λ on INtest, and Section IV-C2/Table II reports the best λ=1 result on the same INtest in support of the abstract's 3.30% claim. Because Table II's GIFDL rows are exactly Table I's λ=1 rows, the headline number is an in-sample selected optimum rather than a held-out prediction. The paper also never reports the threshold τ from Eq. (1), and the section titled 'The optimal value of the parameter τ' actually tunes λ, leaving the fluctuation mechanism underdetermined; the conclusion's own admission that fluctuation differences also appear in smooth background regions further weakens the mechanism story. These are correctness and evidence concerns, not circularity. The self-citation to prior volatility-cost work [47] by overlapping authors is used for the VC baseline and fixed β=0.15, but it is not load-bearing for GIFDL's central claim, so it does not raise the circularity score further.

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

The central claim rests mainly on the empirical premise that fluctuation statistics are a good model for secure modifications, together with a handful of training hyperparameters. No new physical or mathematical entities are introduced.

free parameters (4)
  • λ (adversarial loss weight) = 1
    Weights the D2 fluctuation term in Eq. (12); tuned on the INtest test set in Section IV-C1 and fixed to 1 for all reported results.
  • τ (MSE threshold) = not reported
    Filters fluctuation images in Eq. (1); crucial for security (Table V), but its value is never given.
  • γ (double-tanh steepness) = 60
    Controls the differentiable approximation of ternary embedding in Eq. (5); taken from UT-GAN [7].
  • α, β (generator loss weights) = α=1, β=1e-7
    Set as in UT-GAN [7] for Eq. (11); not fitted in this paper but still hyperparameters of the method.
assumptions (3)
  • domain assumption Fluctuation images model the distribution of secure modifications.
    The method's premise, established in Section III-A and used in the two-discriminator loss, is that matching fluctuation-image statistics improves steganographic security.
  • ad hoc to paper The MSE threshold τ selects useful fluctuation images without losing the signal.
    Eq. (1) filters images, and the ablation (Table V) shows the threshold is necessary, but the criterion is specific to this paper and its value is not specified.
  • domain assumption Alternating updates of the weaker discriminator avoid training collapse.
    Adopted from GMAN [8] in Algorithm 1; the paper relies on this to keep both discriminators effective.

how reviews work

0 comments
Cite this review

Pith. "Pith review of GIFDL: Generated Image Fluctuation Distortion Learning for Enhancing Steganographic Security." pith.science (2026). https://pith.science/paper/TARPMBEY

@misc{pith2026250415139,
  author       = {Pith},
  title        = {Pith review of: GIFDL: Generated Image Fluctuation Distortion Learning for Enhancing Steganographic Security},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/TARPMBEY}},
  note         = {Machine review of arXiv:2504.15139}
}
read the original abstract

Minimum distortion steganography is currently the mainstream method for modification-based steganography. A key issue in this method is how to define steganographic distortion. With the rapid development of deep learning technology, the definition of distortion has evolved from manual design to deep learning design. Concurrently, rapid advancements in image generation have made generated images viable as cover media. However, existing distortion design methods based on machine learning do not fully leverage the advantages of generated cover media, resulting in suboptimal security performance. To address this issue, we propose GIFDL (Generated Image Fluctuation Distortion Learning), a steganographic distortion learning method based on the fluctuations in generated images. Inspired by the idea of natural steganography, we take a series of highly similar fluctuation images as the input to the steganographic distortion generator and introduce a new GAN training strategy to disguise stego images as fluctuation images. Experimental results demonstrate that GIFDL, compared with state-of-the-art GAN-based distortion learning methods, exhibits superior resistance to steganalysis, increasing the detection error rates by an average of 3.30% across three steganalyzers.

Figures

Figures reproduced from arXiv: 2504.15139 by the authors.

Figure 1
Figure 1. Comparison of existing GAN-based steganography methods with the [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. (a) Users can access the black-box Stable Diffusion at [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. (a) Cover image, (b) average pixel difference between the 10 [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figures from the paper (5 more)
Figure 4
Figure 4. Figure 4: The overall framework of the proposed GIFDL consists of three components: the Generative Model, the Generator, and the Discriminator. In the [PITH_FULL_IMAGE:figures/full_fig_p005_4.png]
Figure 5
Figure 5. Figure 5: (a) Image generated with CFG = 7.5000. (b) Image generated with [PITH_FULL_IMAGE:figures/full_fig_p005_5.png]
Figure 6
Figure 6. Figure 6: Cover images (generated by Stable Diffusion) and modification maps with different number of iterations. [PITH_FULL_IMAGE:figures/full_fig_p008_6.png]
Figure 7
Figure 7. Figure 7: Cover images, stego images (first row, third row), and corresponding modification maps (second row, fourth row). [PITH_FULL_IMAGE:figures/full_fig_p009_7.png]
Figure 8
Figure 8. Figure 8: Differences between datasets IN, DDB and MDB. (a) The prompts used in IN come from class labels of ImageNet, and the images in IN are usually single objects with simpler textures. (b) The prompts used in DDB and MDB are from real users, and the images in DDB are more a…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

51 extracted references · 35 canonical work pages

  1. [1]

    Designing steganographic distortion using directional filters,

    V . Holub and J. Fridrich, “Designing steganographic distortion using directional filters,” in 2012 IEEE International workshop on information forensics and security (WIFS) . IEEE, 2012, pp. 234–239

    work page 2012

  2. [2]

    Universal distortion function for steganography in an arbitrary domain,

    V . Holub, J. Fridrich, and T. Denemark, “Universal distortion function for steganography in an arbitrary domain,” EURASIP Journal on Infor- mation Security, vol. 2014, pp. 1–13, 2014

    work page 2014

  3. [3]

    A new cost function for spatial image steganography,

    B. Li, M. Wang, J. Huang, and X. Li, “A new cost function for spatial image steganography,” in 2014 IEEE International conference on image processing (ICIP). IEEE, 2014, pp. 4206–4210. SUBMITTED TO IEEE TIFS 13

    work page 2014

  4. [4]

    Content-adaptive steganog- raphy by minimizing statistical detectability,

    V . Sedighi, R. Cogranne, and J. Fridrich, “Content-adaptive steganog- raphy by minimizing statistical detectability,” IEEE Transactions on Information Forensics and Security , vol. 11, no. 2, pp. 221–234, 2015

    work page 2015

  5. [5]

    Using statistical image model for jpeg steganography: Uniform embedding revisited,

    L. Guo, J. Ni, W. Su, C. Tang, and Y .-Q. Shi, “Using statistical image model for jpeg steganography: Uniform embedding revisited,” IEEE Transactions on Information Forensics and Security , vol. 10, no. 12, pp. 2669–2680, 2015

    work page 2015

  6. [6]

    Automatic steganographic distortion learning using a generative adversarial network,

    W. Tang, S. Tan, B. Li, and J. Huang, “Automatic steganographic distortion learning using a generative adversarial network,” IEEE Signal Processing Letters, vol. 24, no. 10, pp. 1547–1551, 2017

    work page 2017

  7. [7]

    An embedding cost learning framework using gan,

    J. Yang, D. Ruan, J. Huang, X. Kang, and Y .-Q. Shi, “An embedding cost learning framework using gan,” IEEE Transactions on Information Forensics and Security, vol. 15, pp. 839–851, 2019

    work page 2019

  8. [8]

    Steganography embedding cost learning with generative multi-adversarial network,

    D. Huang, W. Luo, M. Liu, W. Tang, and J. Huang, “Steganography embedding cost learning with generative multi-adversarial network,” IEEE Transactions on Information Forensics and Security , 2023

    work page 2023

Show all 51 references
  1. [9]

    Channel attention image steganography with generative adversarial networks,

    J. Tan, X. Liao, J. Liu, Y . Cao, and H. Jiang, “Channel attention image steganography with generative adversarial networks,” IEEE transactions on network science and engineering , vol. 9, no. 2, pp. 888–903, 2021

    2021

  2. [10]

    A robust coverless im- age steganography based on an end-to-end hash generation model,

    L. Meng, X. Jiang, Z. Zhang, Z. Li, and T. Sun, “A robust coverless im- age steganography based on an end-to-end hash generation model,”IEEE Transactions on Circuits and Systems for Video Technology , vol. 33, pp. 3542–3558, 2023. [Online]. Available: https://api.semanticschol...

    2023

  3. [11]

    Lightweight and effective deep image steganalysis network,

    S. Weng, M. Chen, L. Yu, and S. Sun, “Lightweight and effective deep image steganalysis network,” IEEE Signal Processing Letters , vol. 29, pp. 1888–1892, 2022

    2022

  4. [12]

    High- resolution image synthesis with latent diffusion models,

    R. Rombach, A. Blattmann, D. Lorenz, P. Esser, and B. Ommer, “High- resolution image synthesis with latent diffusion models,” in Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2022, pp. 10 684–10 695

    2022

  5. [13]

    Steganogan: High capacity image steganography with gans,

    K. A. Zhang, A. Cuesta-Infante, L. Xu, and K. Veeramachaneni, “Steganogan: High capacity image steganography with gans,” arXiv preprint arXiv:1901.03892, 2019

    1901 arXiv

  6. [14]

    Minimizing additive distortion in steganography using syndrome-trellis codes,

    T. Filler, J. Judas, and J. Fridrich, “Minimizing additive distortion in steganography using syndrome-trellis codes,” IEEE Transactions on Information Forensics and Security , vol. 6, no. 3, pp. 920–935, 2011

    2011

  7. [15]

    Fast and effective global covariance pooling network for image steganalysis,

    X. Deng, B. Chen, W. Luo, and D. Luo, “Fast and effective global covariance pooling network for image steganalysis,” in Proceedings of the ACM workshop on information hiding and multimedia security, 2019, pp. 230–234

    2019

  8. [16]

    Image disentanglement autoencoder for steganography without embedding,

    X. Liu, Z. Ma, J. Ma, J. Zhang, G. Schaefer, and H. Fang, “Image disentanglement autoencoder for steganography without embedding,” in Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2022, pp. 2303–2312

    2022

  9. [17]

    Generative steganography network,

    P. Wei, S. Li, X. Zhang, G. Luo, Z. Qian, and Q. Zhou, “Generative steganography network,” in Proceedings of the 30th ACM International Conference on Multimedia , 2022, pp. 1621–1629

    2022

  10. [18]

    Stegad- dpm: Generative image steganography based on denoising diffusion probabilistic model,

    Y . Peng, D. Hu, Y . Wang, K. Chen, G. Pei, and W. Zhang, “Stegad- dpm: Generative image steganography based on denoising diffusion probabilistic model,” in Proceedings of the 31st ACM International Conference on Multimedia , 2023, pp. 7143–7151

    2023

  11. [19]

    Yedroudj-net: An efficient cnn for spatial steganalysis,

    M. Yedroudj, F. Comby, and M. Chaumont, “Yedroudj-net: An efficient cnn for spatial steganalysis,” in 2018 IEEE international conference on acoustics, speech and signal processing (ICASSP) . IEEE, 2018, pp. 2092–2096

    2018

  12. [20]

    On the limits of steganography,

    R. J. Anderson and F. A. Petitcolas, “On the limits of steganography,” IEEE Journal on selected areas in communications , vol. 16, no. 4, pp. 474–481, 1998

    1998

  13. [21]

    Practical steganalysis of digital images: state of the art,

    J. Fridrich and M. Goljan, “Practical steganalysis of digital images: state of the art,” security and Watermarking of Multimedia Contents IV , vol. 4675, pp. 1–13, 2002

    2002

  14. [22]

    Lsb matching revisited,

    J. Mielikainen, “Lsb matching revisited,” IEEE signal processing letters, vol. 13, no. 5, pp. 285–287, 2006

    2006

  15. [23]

    Gibbs construction in steganography,

    T. Filler and J. Fridrich, “Gibbs construction in steganography,” IEEE Transactions on Information Forensics and Security , vol. 5, no. 4, pp. 705–720, 2010

    2010

  16. [24]

    Optimality of polar codes in additive steganography under constant distortion profile,

    Q. Yao, W. Zhang, and N. Yu, “Optimality of polar codes in additive steganography under constant distortion profile,” in 2022 14th Interna- tional Conference on Wireless Communications and Signal Processing (WCSP). IEEE, 2022, pp. 404–408

    2022

  17. [25]

    Distribution-preserving steganography based on text-to-speech genera- tive models,

    K. Chen, H. Zhou, H. Zhao, D. Chen, W. Zhang, and N. Yu, “Distribution-preserving steganography based on text-to-speech genera- tive models,” IEEE Transactions on Dependable and Secure Computing, vol. 19, no. 5, pp. 3343–3356, 2021

    2021

  18. [26]

    Structural design of convolutional neural networks for steganalysis,

    G. Xu, H.-Z. Wu, and Y .-Q. Shi, “Structural design of convolutional neural networks for steganalysis,” IEEE Signal Processing Letters , vol. 23, no. 5, pp. 708–712, 2016

    2016

  19. [27]

    Denoising diffusion probabilistic models,

    J. Ho, A. Jain, and P. Abbeel, “Denoising diffusion probabilistic models,” Advances in neural information processing systems , vol. 33, pp. 6840– 6851, 2020

    2020

  20. [28]

    Generative adversarial text to image synthesis,

    S. Reed, Z. Akata, X. Yan, L. Logeswaran, B. Schiele, and H. Lee, “Generative adversarial text to image synthesis,” in International con- ference on machine learning . PMLR, 2016, pp. 1060–1069

    2016

  21. [29]

    Stackgan: Text to photo-realistic image synthesis with stacked generative adversarial networks,

    H. Zhang, T. Xu, H. Li, S. Zhang, X. Wang, X. Huang, and D. N. Metaxas, “Stackgan: Text to photo-realistic image synthesis with stacked generative adversarial networks,” in Proceedings of the IEEE interna- tional conference on computer vision , 2017, pp. 5907–5915

    2017

  22. [30]

    Attngan: Fine-grained text to image generation with attentional generative adversarial networks,

    T. Xu, P. Zhang, Q. Huang, H. Zhang, Z. Gan, X. Huang, and X. He, “Attngan: Fine-grained text to image generation with attentional generative adversarial networks,” in Proceedings of the IEEE conference on computer vision and pattern recognition , 2018, pp. 1316–1324

    2018

  23. [31]

    Controllable text-to- image generation,

    B. Li, X. Qi, T. Lukasiewicz, and P. Torr, “Controllable text-to- image generation,” Advances in Neural Information Processing Systems, vol. 32, 2019

    2019

  24. [32]

    Zero-shot text-to-image generation,

    A. Ramesh, M. Pavlov, G. Goh, S. Gray, C. V oss, A. Radford, M. Chen, and I. Sutskever, “Zero-shot text-to-image generation,” in International Conference on Machine Learning . PMLR, 2021, pp. 8821–8831

    2021

  25. [33]

    Gpt-4 technical report,

    OpenAI, “Gpt-4 technical report,” in Proceedings of the Confer- ence on Artificial Intelligence , 2023. [Online]. Available: https://api. semanticscholar.org/CorpusID:257532815

    2023

  26. [34]

    Photorealistic text-to-image diffusion models with deep language understanding,

    C. Saharia, W. Chan, S. Saxena, L. Li, J. Whang, E. L. Denton, K. Ghasemipour, R. Gontijo Lopes, B. Karagol Ayan, T. Salimans et al., “Photorealistic text-to-image diffusion models with deep language understanding,” Advances in Neural Information Processing Systems , vol. 35, ...

    2022

  27. [35]

    Attention based data hiding with generative adversarial net- works,

    C. Yu, “Attention based data hiding with generative adversarial net- works,” in Proceedings of the AAAI conference on artificial intelligence , vol. 34, no. 01, 2020, pp. 1120–1128

    2020

  28. [36]

    High-capacity convolutional video steganography with temporal residual modeling,

    X. Weng, Y . Li, L. Chi, and Y . Mu, “High-capacity convolutional video steganography with temporal residual modeling,” in Proceedings of the 2019 on international conference on multimedia retrieval, 2019, pp. 87– 95

    2019

  29. [38]

    Deepmih: Deep invertible network for multiple image hiding,

    Z. Guan, J. Jing, X. Deng, M. Xu, L. Jiang, Z. Zhang, and Y . Li, “Deepmih: Deep invertible network for multiple image hiding,” IEEE Transactions on Pattern Analysis and Machine Intelligence , vol. 45, pp. 372–390, 2022. [Online]. Available: https://api.semanticscholar.org/ Co...

    2022

  30. [39]

    Imagenet: A large-scale hierarchical image database,

    J. Deng, W. Dong, R. Socher, L.-J. Li, K. Li, and L. Fei-Fei, “Imagenet: A large-scale hierarchical image database,” in 2009 IEEE Conference on Computer Vision and Pattern Recognition , 2009, pp. 248–255

    2009

  31. [40]

    Deep residual network for steganalysis of digital images,

    M. Boroumand, M. Chen, and J. Fridrich, “Deep residual network for steganalysis of digital images,” IEEE Transactions on Information Forensics and Security, vol. 14, no. 5, pp. 1181–1193, 2018

    2018

  32. [41]

    Improving image generation with better captions,

    J. Betker, G. Goh, L. Jing, T. Brooks, J. Wang, L. Li, L. Ouyang, J. Zhuang, J. Lee, Y . Guo et al. , “Improving image generation with better captions,” 2023

    2023

  33. [42]

    Generative adversarial networks,

    I. Goodfellow, J. Pouget-Abadie, M. Mirza, B. Xu, D. Warde-Farley, S. Ozair, A. Courville, and Y . Bengio, “Generative adversarial networks,” Communications of the ACM , vol. 63, no. 11, pp. 139–144, 2020

    2020

  34. [43]

    Rich models for steganalysis of digital images,

    J. Fridrich and J. Kodovsky, “Rich models for steganalysis of digital images,” IEEE Transactions on information Forensics and Security , vol. 7, no. 3, pp. 868–882, 2012

    2012

  35. [44]

    Defining cost functions for adaptive jpeg steganography at the microscale,

    K. Chen, H. Zhou, W. Zhou, W. Zhang, and N. Yu, “Defining cost functions for adaptive jpeg steganography at the microscale,” IEEE Transactions on Information Forensics and Security , vol. 14, no. 4, pp. 1052–1066, 2018

    2018

  36. [45]

    Ldgm codes based near- optimal coding for adaptive steganography,

    Q. Yao, W. Zhang, K. Chen, and N. Yu, “Ldgm codes based near- optimal coding for adaptive steganography,” IEEE Transactions on Communications, 2023

    2023

  37. [46]

    Steganography via cover-source switching,

    P. Bas, “Steganography via cover-source switching,” in 2016 IEEE International Workshop on Information Forensics and Security (WIFS) . IEEE, 2016, pp. 1–6

    2016

  38. [47]

    Steganography with generated images: Leveraging volatility to enhance security,

    J. Zhang, K. Chen, W. Li, W. Zhang, and N. Yu, “Steganography with generated images: Leveraging volatility to enhance security,” IEEE Transactions on Dependable and Secure Computing , 2023. SUBMITTED TO IEEE TIFS 14

    2023

  39. [48]

    Diffusiondb: A large-scale prompt gallery dataset for text-to- image generative models,

    Z. J. Wang, E. Montoya, D. Munechika, H. Yang, B. Hoover, and D. H. Chau, “Diffusiondb: A large-scale prompt gallery dataset for text-to- image generative models,” arXiv preprint arXiv:2210.14896 , 2022

    2022 arXiv

  40. [49]

    A steganography immunoprocess- ing framework against cnn-based and handcrafted steganalysis,

    Y . Chen, H. Wang, W. Li, and W. Li, “A steganography immunoprocess- ing framework against cnn-based and handcrafted steganalysis,” IEEE Transactions on Information Forensics and Security , 2024

    2024

  41. [50]

    Genimage: A million-scale benchmark for detecting ai- generated image,

    M. Zhu, H. Chen, Q. Yan, X. Huang, G. Lin, W. Li, Z. Tu, H. Hu, J. Hu, and Y . Wang, “Genimage: A million-scale benchmark for detecting ai- generated image,” 2023

    2023

  42. [51]

    Gan-based adaptive cost learning for enhanced image steganography security,

    D. Wang, G. Yang, J. Chen, and X. Ding, “Gan-based adaptive cost learning for enhanced image steganography security,” Expert Systems with Applications, vol. 249, p. 123471, 2024

    2024

  43. [52]

    Journeydb: A benchmark for generative image understanding,

    K. Sun, J. Pan, Y . Ge, H. Li, H. Duan, X. Wu, R. Zhang, A. Zhou, Z. Qin, Y . Wang et al., “Journeydb: A benchmark for generative image understanding,” Advances in neural information processing systems , vol. 36, pp. 49 659–49 678, 2023

    2023

Pith tools

Reviewed August 16, 2026 · model on record in the stance chip above.