Pith. sign in

REVIEW 3 major objections 6 minor 47 references

MLEP: Multi-granularity Local Entropy Patterns for Universal AI-generated Image Detection

T0 review · 3 major / 6 minor · reviewed 2026-08-16 · deepseek-v4-flash

Pith's one-line read Local entropy patterns spot AI images across 32 generators

desk verdict A plausible, cheap entropy-based feature for AIGI detection with a real but possibly inflated mean gain: the headline 97.1% is selected from ablations on the same test datasets, with no held-out validation or error bars. read the letter →

arxiv 2504.13726 v2 pith:WUKY4Z5X submitted 2025-04-18 cs.CV

classification cs.CV
keywords AI-generatedimagedetectionlocalentropypatchshufflingmulti-scaleresamplingShannondeepfakecross-domaingeneralizationProGAN
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

Generative image models leave subtle statistical traces even when pixel-level artifacts are invisible. This paper proposes that the right trace is local information entropy: how randomly arranged pixel values are inside tiny 2x2 windows, measured after the image has been scrambled into small shuffled patches and resampled at several scales. It claims this Multi-granularity Local Entropy Pattern (MLEP) is content-agnostic, so a detector trained on images from one GAN, ProGAN, can recognize images from 31 other generators, including diffusion models. On its 32-dataset benchmark the detector reaches 97.1% mean accuracy and 98.9% average precision, outperforming prior pixel-difference and pretrained-model baselines. If the claim holds, entropy maps give a cheap, source-invariant cue for deepfake and misinformation defenses.

What carries the argument

The load-bearing object is the MLEP feature map. Construction: split an image into $2\times2$ patches, shuffle them randomly (per color channel), downsample and upsample the scrambled image at scales $S=\{1,1/2,1/4\}$ with bilinear interpolation, then slide a $2\times2$ window with stride 1 over every resampled version and record the Shannon entropy of the four pixels inside each window. Because only four values enter the entropy formula, the LEP output takes values in $\{0,0.8,1.0,1.5,2.0\}$, and the overlapping windows capture intra-patch, inter-patch, and inter-scale randomness at once. Patch shuffling destroys the semantic layout, multi-scale resampling exposes upsampling artifacts common to generators, and the entropy statistic replaces pixel contrasts with a measure of local randomness. The resulting multi-channel maps are concatenated and fed to a standard ResNet-50 trained with binary cross-entropy.

What would settle it

Fix hyperparameters ($l=2$, $S=\{1,1/2,1/4\}$, stride 1, bilinear interpolation) using a held-out validation set of a few generators, then test on a generator released after training that is not among the 32, or apply mild JPEG compression and Gaussian blur to the existing test sets. If mean accuracy drops to baseline levels, the claim of a universal entropy fingerprint fails.

Watch

Extended reading notes

Core claim

The paper's central discovery is that local entropy distributions over shuffled micro-patches separate real photographs from AI-generated images in a way that transfers across generator families. A real image consistently has a higher probability of maximum entropy (2.0) in 2x2 windows than a generated one, which the authors attribute to the glossy, over-smoothed texture generators produce. By computing entropy on randomly permuted $2\times2$ patches across the resampling scales $S=\{1,1/2,1/4\}$, MLEP suppresses scene semantics while preserving and amplifying generator-specific randomness. Trained only on ProGAN images from ForenSynths and real LSUN images, a ResNet-50 classifier on MLEP maps achieves 97.1% mean accuracy over 32 generation models, with gains of 3.1% in accuracy and 1.6% in average precision over the nearest strong baseline NPR. The authors frame this as the first use of image entropy as a detection cue for AI-generated images.

Load-bearing premise

The entire reported advantage rests on the assumption that the hyperparameters were not effectively tuned to the 32 test datasets; the paper selects patch size, scales, stride, and interpolation through ablations on those same datasets without describing a held-out validation split for model selection.

Editorial extensions

If this is right

  • A detector trained only on ProGAN images transfers to 16 GAN and 16 diffusion models, reaching 97.1% mean accuracy across all 32.
  • Patch size $l=2$ beats $l=4$ and $l=8$, supporting the claim that finer semantic scrambling improves content-agnostic detection.
  • Entropy alone, without shuffling or multi-scale resampling, already exceeds the NPR baseline, indicating the entropy statistic itself carries the signal.
  • Multi-scale fusion with scales $\{1,1/2,1/4\}$ is optimal; adding $1/8$ hurts, so moderate resampling introduces useful artifacts without excessive distortion.
  • Performance is stable across ResNet-18, ResNet-34, ResNet-50, and ResNet-101, so the representation rather than a particular classifier depth drives the result.

Reading between the lines

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

  • One testable extension: evaluate MLEP under JPEG compression, blur, and noise; the paper's own limitation section says these are unstudied, and a drop there would bound its real-world use.
  • Because the feature is a five-value entropy pattern, a fast implementation could classify images in real time on edge devices; the paper does not measure latency or throughput.
  • If over-smoothed texture is the mechanism, MLEP should also catch partially generated edits such as face swaps and inpainting, a benchmark the paper does not include.
  • The claim of content agnosticism could be stress-tested by training on LSUN scenes and testing on face or document image datasets, where texture statistics differ.
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 / 6 minor

Summary. The paper proposes Multi-granularity Local Entropy Patterns (MLEP), a handcrafted feature representation for AI-generated image detection. The method computes Shannon entropy over 2x2 sliding windows on randomly shuffled small patches, repeats this at multiple resampling scales, and feeds the resulting multi-channel entropy maps to a ResNet classifier trained only on ProGAN images. The authors evaluate on 32 datasets from 16 GAN and 16 diffusion models and report a mean accuracy of 97.1%, surpassing the NPR baseline by 3.1 accuracy points. They also provide ablations on patch size, scale factors, interpolation method, stride, and backbone, plus qualitative visualizations of entropy maps and t-SNE features to support their semantic-suppression argument.

Significance. If the reported results are unbiased, MLEP is an interesting and practically simple contribution: it requires no pretrained generative model or vision-language backbone, it uses only a standard CNN on a derived entropy representation, and its cross-generator generalization (trained on ProGAN, tested on 32 generators) is strong. The paper also gives a clear qualitative story for why patch shuffling and local entropy should suppress semantic content. These strengths are real and make the method worth pursuing. However, the central quantitative claim — a 3.1-point mean accuracy advantage over NPR and a 5.2-point advantage over FatFormer — is currently supported by an evaluation protocol that selects hyperparameters on the same test datasets that produce the headline numbers, and it reports no error bars or significance tests. The claim of 'universal' detection is also weakened by per-dataset results where the method is far behind the baselines. The significance of the paper as written therefore rests on an evaluation that needs to be made statistically sound before the conclusions can be trusted.

major comments (3)
  1. [§4.1 and §4.2.2 (Tables 5–8)] The hyperparameters of MLEP — patch size l=2, scales S={1,1/2,1/4}, bilinear interpolation, stride 1, and the ResNet-50 backbone — are selected by maximizing mean accuracy/AP on the same 16 GAN and 16 diffusion test datasets that are later used for the headline results in Tables 1–3. The text does not describe any held-out validation split. This is selection on the test distribution and can inflate the reported 97.1% mean accuracy. It also makes the comparison to NPR and FatFormer unfair, since those baselines were not tuned on these 32 datasets and several were taken directly from the literature with their published settings. Please introduce a proper validation split (e.g., held-out categories of ForenSynths, or a separate set of generative models for validation), or report results for all hyperparameter combinations and discuss sensitivity explicitly.
  2. [§4.2.1 (Tables 1–3)] All reported numbers come from a single run, and the seed for the random patch shuffling is not reported. Because patch shuffling introduces randomness, the mean accuracy and AP could vary from run to run; the claimed advantage over NPR (97.1 vs 94.0) could be within run-to-run noise. Please report means and standard deviations over at least five runs for the key comparisons, and include a paired significance test (e.g., a paired t-test or Wilcoxon test over the 32 datasets) against the strongest baselines NPR and FatFormer.
  3. [§4.2.1, Tables 1 and 3] The paper concludes that MLEP 'outperforms state-of-the-art approaches by over 3% in accuracy' and calls the method 'universal', but this is based only on the mean over all 32 datasets. On the GAN subset, MLEP achieves 96.4 Acc versus FatFormer's 99.0 (Table 3), and on several individual GAN datasets the gap is much larger: BigGAN 87.1 vs 99.5 for FatFormer, GauGAN 82.0 vs 99.4, and S3GAN 83.4 vs 99.0 (Table 1). These are not isolated drops; they are on 3 of the 16 GAN datasets and contradict the impression of uniform superiority. The authors should report per-domain results, discuss the GAN-specific weakness, and temper the conclusion accordingly. A mean-only reporting is not sufficient to substantiate the 'universal' claim.
minor comments (6)
  1. [Abstract] "multiple image scaled" should be "multiple image scales".
  2. [§3.3] The set V={0, 0.8, 1.0, 1.5, 2} suggests that entropy values are rounded to one decimal; please define the rounding rule explicitly and state how ties in the four pixel values are handled in the probability estimate.
  3. [§3.3] The text says "The proof and an efficient computation algorithm for LEP on a 2×2 window are provided in the supplementary material." If the supplementary material is not part of this submission, the proof should be included in the main text or omitted; otherwise the statement is not verifiable.
  4. [Figure 1] The caption lists percentages (e.g., 45.6%, 43.9%) but does not explain what they represent; please clarify whether they are the fraction of pixels with entropy 2.0, or some other quantity.
  5. [References] Reference [1] has a typo: "Cramer Cistance" should be "Cramer Distance", and reference [27] has "Fiffusion" which should be "Diffusion".
  6. [§4.2.3] The t-SNE visualizations (Fig. 6) are qualitative; consider adding a quantitative separation metric (e.g., silhouette score or nearest-neighbor accuracy) so the claim of cleaner separation is testable.

Circularity Check

1 steps flagged · score 6.0 of 10

The headline 97.1% mean accuracy is selected on the same test datasets, so the reported 'prediction' is partly a fitted maximum rather than an unbiased generalization estimate.

  1. fitted input called prediction [Section 4.1 Implementation Details; Section 4.2.2 Table 5; final results in Section 4.2.1 Tables 1-3]
    "To evaluate the effects of patch size (l) and resampling scales (S), we tested l∈{2, 4, 8} and S ={1, 1/2, 1/4, 1/8}. ... The best performance was achieved with the smallest patch size (l = 2), indicating that stronger semantic scrambling improves detection. ... We evaluated the generalization performance ... using patch size l = 2 , scales S = {1, 1/2, 1/4}, and a ResNet-50 backbone."

    Tables 5-8 report Acc./A.P. on the same GAN-set, Diff.-set, and Mean collections that later produce the headline numbers in Tables 1-3. The final configuration (l=2, S={1,1/2,1/4}, stride 1, bilinear, ResNet-50) is chosen because it achieved the best performance in Table 5, then the same test datasets are used to report 97.1% mean accuracy. Thus the reported number is the maximum over a small hyperparameter grid evaluated with test labels, not the expected accuracy of a pre-specified method. The paper describes no held-out validation split, and baselines are not reported as equivalently tuned on these 32 datasets, so the 3.1-point margin over NPR is partly a selection artifact.

full rationale

The feature-construction part of the paper is not circular: MLEP is explicitly defined via patch shuffling, multi-scale resampling, and Shannon entropy on 2x2 windows, and the classifier is trained only on ProGAN images. There is no self-citation chain, no imported uniqueness theorem, and no renaming of a known result; the entropy descriptor is self-contained and the empirical comparison is meaningful in principle. The circularity is confined to the evaluation protocol: hyperparameters are selected using the same test datasets whose accuracy is then reported as the method's generalization performance. This makes the headline 97.1% figure, and the claimed >3% gain over NPR, partly constructed from the test labels rather than predicted. Because the central efficacy claim depends on that reported number, the paper earns a partial-circularity score of 6 rather than a clean bill of health. If the authors re-ran the protocol with a held-out validation split or pre-registered hyperparameters, the same method could receive a low circularity score.

Assumptions & free parameters 4 free parameters · 5 assumptions · 0 invented entities

The method's central claim rests on several domain assumptions about entropy, shuffling, and resampling artifacts that are tested only empirically on the chosen datasets. The hyperparameters are hand-picked via ablation on the test data, which is a form of tuning. No new theoretical entities are introduced; the only 'new thing' is the feature representation and the empirical claim.

free parameters (4)
  • patch_size_l = 2
    Selected by ablation on the same test datasets (Table 5); smaller patches give higher accuracy, but l=1 was not tested.
  • resampling_scales_S = {1, 1/2, 1/4}
    Selected by ablation (Table 5); adding 1/8 reduces accuracy, removing 1/4 also reduces it.
  • sliding_window_stride = 1
    Stride 1 outperforms stride 2 (Table 7).
  • interpolation_method = bilinear
    Bilinear and bicubic are comparable; bilinear selected (Table 6).
assumptions (5)
  • domain assumption Shannon entropy on a 2x2 window, with values confined to {0, 0.8, 1.0, 1.5, 2}, measures pixel randomness relevant to generation artifacts
    Section 3.3 and Fig. 1. The paper assumes this statistic is discriminative and sufficient; this is the core hypothesis, validated only empirically on the chosen datasets.
  • domain assumption Random shuffling of small patches removes semantic content without destroying the artifacts entropy is meant to capture
    Section 3.1. The claim that fine-grained shuffling (l=2) suppresses semantics better than 32x32 shuffling [45] is asserted from qualitative inspection, not proven.
  • domain assumption Downsampling and upsampling the scrambled image introduces resampling artifacts that differ between real and generated images
    Section 3.2. The hypothesis is inspired by [38] and supported only by the ablation gain.
  • domain assumption A ResNet-50 CNN trained on MLEP maps of ProGAN images can generalize to unseen GANs and diffusion models
    Section 4.1. The transfer is the core generalization claim, tested on 32 datasets but with no independent validation split.
  • standard math Standard arithmetic and information theory definitions
    Entropy formula and arithmetic are standard; no novel math is used.

how reviews work

0 comments
Cite this review

Pith. "Pith review of MLEP: Multi-granularity Local Entropy Patterns for Universal AI-generated Image Detection." pith.science (2026). https://pith.science/paper/WUKY4Z5X

@misc{pith2026250413726,
  author       = {Pith},
  title        = {Pith review of: MLEP: Multi-granularity Local Entropy Patterns for Universal AI-generated Image Detection},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/WUKY4Z5X}},
  note         = {Machine review of arXiv:2504.13726}
}
read the original abstract

Advancements in image generation technologies have raised significant concerns about their potential misuse, such as producing misinformation and deepfakes. Therefore, there is an urgent need for effective methods to detect AI-generated images (AIGI). Despite progress in AIGI detection, achieving reliable performance across diverse generation models and scenes remains challenging due to the lack of source-invariant features and limited generalization capabilities in existing methods. In this work, we explore the potential of using image entropy as a cue for AIGI detection and propose Multi-granularity Local Entropy Patterns (MLEP), a set of entropy feature maps computed across shuffled small patches over multiple image scaled. MLEP comprehensively captures pixel relationships across dimensions and scales while significantly disrupting image semantics, reducing potential content bias. Leveraging MLEP, a robust CNN-based classifier for AIGI detection can be trained. Extensive experiments conducted in an open-world scenario, evaluating images synthesized by 32 distinct generative models, demonstrate significant improvements over state-of-the-art methods in both accuracy and generalization.

Figures

Figures reproduced from arXiv: 2504.13726 by the authors.

Figure 1
Figure 1. Comparison of local entropy distributions between real and AI-generated images using [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. The MLEP framework, composed of three steps: Patch Shuffling, Multi-Scale Resampling, [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. Illustration of LEP computation in a single patch (a) and its intra-block and inter-block [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figures from the paper (3 more)
Figure 4
Figure 4. Figure 4: Visualization of local entropy patterns for several real–fake image pairs, along with their [PITH_FULL_IMAGE:figures/full_fig_p008_4.png]
Figure 5
Figure 5. Figure 5: Qualitative comparison among Zheng [45], NPR [38], and our method. LEP preserves minimal visible semantics, while MLEP (without resampling) further suppresses semantic content. comparing our method with two competitive baselines—NPR [38] and FreqNet [37]—which also use…
Figure 6
Figure 6. Figure 6: t-SNE visualization of real vs. fake samples across [PITH_FULL_IMAGE:figures/full_fig_p009_6.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

47 extracted references · 40 canonical work pages

  1. [1]

    The Cramer Cistance as a Solution to Biased Wasserstein Gradients

    Marc G Bellemare, Ivo Danihelka, Will Dabney, Shakir Mohamed, Balaji Lakshminarayanan, Stephan Hoyer, and Rémi Munos. The Cramer Cistance as a Solution to Biased Wasserstein Gradients. In International Conference on Learning Representations, 2018

  2. [2]

    BEGAN: Boundary Equilibrium Generative Adversarial Networks

    David Berthelot. BEGAN: Boundary Equilibrium Generative Adversarial Networks. arXiv preprint arXiv:1703.10717, 2017

  3. [3]

    Large Scale GAN Training for High Fidelity Natural Image Synthesis

    Andrew Brock, Jeff Donahue, and Karen Simonyan. Large Scale GAN Training for High Fidelity Natural Image Synthesis. In International Conference on Learning Representations, 2018

  4. [4]

    DRCT: Diffusion Reconstruction Contrastive Training towards Universal Detection of Diffusion Generated Images

    Baoying Chen, Jishen Zeng, Jianquan Yang, and Rui Yang. DRCT: Diffusion Reconstruction Contrastive Training towards Universal Detection of Diffusion Generated Images. In Forty-first International Conference on Machine Learning, 2024

  5. [5]

    StarGAN: Unified Generative Adversarial Networks for Multi-Domain Image-to-Image Transla- tion

    Yunjey Choi, Minje Choi, Munyoung Kim, Jung-Woo Ha, Sunghun Kim, and Jaegul Choo. StarGAN: Unified Generative Adversarial Networks for Multi-Domain Image-to-Image Transla- tion. In Proceedings of the IEEE conference on computer vision and pattern recognition, pages 8789–8797, 2018

  6. [6]

    Diffusion Models Beat GANs on Image Synthesis

    Prafulla Dhariwal and Alexander Nichol. Diffusion Models Beat GANs on Image Synthesis. Advances in neural information processing systems, 34:8780–8794, 2021

  7. [7]

    Leveraging Frequency Analysis for Deep Fake Image Recognition

    Joel Frank, Thorsten Eisenhofer, Lea Schönherr, Asja Fischer, Dorothea Kolossa, and Thorsten Holz. Leveraging Frequency Analysis for Deep Fake Image Recognition. In International Conference on Machine Learning, pages 3247–3258. PMLR, 2020

  8. [8]

    Goodfellow, Jean Pouget-Abadie, Mehdi Mirza, Bing Xu, David Warde-Farley, Sherjil Ozair, Aaron Courville, and Yoshua Bengio

    Ian J. Goodfellow, Jean Pouget-Abadie, Mehdi Mirza, Bing Xu, David Warde-Farley, Sherjil Ozair, Aaron Courville, and Yoshua Bengio. Generative Adversarial Nets. InProceedings of the 28th International Conference on Neural Information Processing Systems - Volume 2, NIPS’14, page 2672–2680, 2014

Show all 47 references
  1. [9]

    Vector Quantized Diffusion Model for Text-to-Image Synthesis

    Shuyang Gu, Dong Chen, Jianmin Bao, Fang Wen, Bo Zhang, Dongdong Chen, Lu Yuan, and Baining Guo. Vector Quantized Diffusion Model for Text-to-Image Synthesis. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 10696–10706, 2022

  2. [10]

    Deep Residual Learning for Image Recognition

    Kaiming He, Xiangyu Zhang, Shaoqing Ren, and Jian Sun. Deep Residual Learning for Image Recognition. In 2016 IEEE Conference on Computer Vision and Pattern Recognition, pages 770–778, 2016

  3. [11]

    AttGAN: Facial Attribute Editing by Only Changing What You Want

    Zhenliang He, Wangmeng Zuo, Meina Kan, Shiguang Shan, and Xilin Chen. AttGAN: Facial Attribute Editing by Only Changing What You Want. IEEE transactions on image processing, 28(11):5464–5478, 2019

  4. [12]

    Denoising Diffusion Probabilistic Models.Advances in neural information processing systems, 33:6840–6851, 2020

    Jonathan Ho, Ajay Jain, and Pieter Abbeel. Denoising Diffusion Probabilistic Models.Advances in neural information processing systems, 33:6840–6851, 2020

  5. [13]

    This isn’t the Nature Podcast-how deepfakes are distorting reality

    Nick Petri´c Howe and Benjamin Thompson. This isn’t the Nature Podcast-how deepfakes are distorting reality. Nature, 2023

  6. [14]

    Progressive Growing of Gans for Improved Quality, Stability, and Variation

    Tero Karras, Timo Aila, Samuli Laine, and Jaakko Lehtinen. Progressive Growing of Gans for Improved Quality, Stability, and Variation. In International Conference on Learning Representations, 2018

  7. [15]

    A Style-Based Generator Architecture for Genera- tive Adversarial Networks

    Tero Karras, Samuli Laine, and Timo Aila. A Style-Based Generator Architecture for Genera- tive Adversarial Networks. In 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 4396–4405, 2019. 10

  8. [16]

    Analyzing and Improving the Image Quality of StyleGAN

    Tero Karras, Samuli Laine, Miika Aittala, Janne Hellsten, Jaakko Lehtinen, and Timo Aila. Analyzing and Improving the Image Quality of StyleGAN. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 8110–8119, 2020

  9. [17]

    CLIPping the Deception: Adapting Vision- Language Models for Universal deepfake detection

    Sohail Ahmed Khan and Duc-Tien Dang-Nguyen. CLIPping the Deception: Adapting Vision- Language Models for Universal deepfake detection. In Proceedings of the 2024 International Conference on Multimedia Retrieval, pages 1006–1015, 2024

  10. [18]

    InfoMax-GAN: Improved Adversarial Image Generation via Information Maximization and Contrastive Learning

    Kwot Sin Lee, Ngoc-Trung Tran, and Ngai-Man Cheung. InfoMax-GAN: Improved Adversarial Image Generation via Information Maximization and Contrastive Learning. In Proceedings of the IEEE/CVF winter conference on applications of computer vision, pages 3942–3952, 2021

  11. [19]

    MMD GAN: Towards Deeper Understanding of Moment Matching Network

    Chun-Liang Li, Wei-Cheng Chang, Yu Cheng, Yiming Yang, and Barnabás Póczos. MMD GAN: Towards Deeper Understanding of Moment Matching Network. Advances in neural information processing systems, 30, 2017

  12. [20]

    Detecting Generated Images by Real Images

    Bo Liu, Fan Yang, Xiuli Bi, Bin Xiao, Weisheng Li, and Xinbo Gao. Detecting Generated Images by Real Images. In European Conference on Computer Vision, pages 95–110. Springer, 2022

  13. [21]

    Forgery-aware Adaptive Transformer for Generalizable Synthetic Image Detection

    Huan Liu, Zichang Tan, Chuangchuang Tan, Yunchao Wei, Jingdong Wang, and Yao Zhao. Forgery-aware Adaptive Transformer for Generalizable Synthetic Image Detection. In Pro- ceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , pages 10770–10780, 2024

  14. [22]

    Pseudo Numerical Methods for Diffusion Models on Manifolds

    Luping Liu, Yi Ren, Zhijie Lin, and Zhou Zhao. Pseudo Numerical Methods for Diffusion Models on Manifolds. In International Conference on Learning Representations, 2022

  15. [23]

    STGAN: A Unified Selective Transfer Network for Arbitrary Image Attribute Editing

    Ming Liu, Yukang Ding, Min Xia, Xiao Liu, Errui Ding, Wangmeng Zuo, and Shilei Wen. STGAN: A Unified Selective Transfer Network for Arbitrary Image Attribute Editing. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 3673–3682, 2019

  16. [24]

    High-Fidelity Image Generation with Fewer Labels

    Mario Luˇci´c, Michael Tschannen, Marvin Ritter, Xiaohua Zhai, Olivier Bachem, and Sylvain Gelly. High-Fidelity Image Generation with Fewer Labels. In International conference on machine learning, pages 4183–4192. PMLR, 2019

  17. [25]

    Generalizing Face Forgery Detection with High-Frequency Features

    Yuchen Luo, Yong Zhang, Junchi Yan, and Wei Liu. Generalizing Face Forgery Detection with High-Frequency Features. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 16317–16326, 2021

  18. [26]

    Spectral Normaliza- tion for Generative Adversarial Networks

    Takeru Miyato, Toshiki Kataoka, Masanori Koyama, and Yuichi Yoshida. Spectral Normaliza- tion for Generative Adversarial Networks. In International Conference on Learning Represen- tations, 2018

  19. [27]

    Improved Denoising Fiffusion Probabilistic Models

    Alexander Quinn Nichol and Prafulla Dhariwal. Improved Denoising Fiffusion Probabilistic Models. In International conference on machine learning, pages 8162–8171. PMLR, 2021

  20. [28]

    Glide: Towards photorealistic image genera- tion and editing with text-guided diffusion models

    Alexander Quinn Nichol, Prafulla Dhariwal, Aditya Ramesh, Pranav Shyam, Pamela Mishkin, Bob Mcgrew, Ilya Sutskever, and Mark Chen. Glide: Towards photorealistic image genera- tion and editing with text-guided diffusion models. In International Conference on Machine Learning, p...

  21. [29]

    RelGAN: Relational Generative Adversarial Networks for Text Generation

    Weili Nie, Nina Narodytska, and Ankit Patel. RelGAN: Relational Generative Adversarial Networks for Text Generation. In International conference on learning representations, 2018

  22. [30]

    Towards Universal Fake Image Detectors that Generalize Across Generative Models

    Utkarsh Ojha, Yuheng Li, and Yong Jae Lee. Towards Universal Fake Image Detectors that Generalize Across Generative Models. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 24480–24489, 2023

  23. [31]

    Semantic Image Synthe- sis with Spatially-Adaptive Normalization

    Taesung Park, Ming-Yu Liu, Ting-Chun Wang, and Jun-Yan Zhu. Semantic Image Synthe- sis with Spatially-Adaptive Normalization. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 2337–2346, 2019. 11

  24. [32]

    Thinking in Frequency: Face Forgery Detection by Mining Frequency-Aware Clues

    Yuyang Qian, Guojun Yin, Lu Sheng, Zixuan Chen, and Jing Shao. Thinking in Frequency: Face Forgery Detection by Mining Frequency-Aware Clues. In European conference on computer vision, pages 86–103. Springer, 2020

  25. [33]

    Zero-Shot Text-to-Image Generation

    Aditya Ramesh, Mikhail Pavlov, Gabriel Goh, Scott Gray, Chelsea V oss, Alec Radford, Mark Chen, and Ilya Sutskever. Zero-Shot Text-to-Image Generation. In International conference on machine learning, pages 8821–8831. PMLR, 2021

  26. [34]

    High- Resolution Image Synthesis with Latent Diffusion Models

    Robin Rombach, Andreas Blattmann, Dominik Lorenz, Patrick Esser, and Björn Ommer. High- Resolution Image Synthesis with Latent Diffusion Models. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 10684–10695, 2022

  27. [35]

    A mathematical theory of communication

    Claude Elwood Shannon. A mathematical theory of communication. The Bell system technical journal, 27(3):379–423, 1948

  28. [36]

    GANGen-Detection: A Dataset Generated by GANs for Generalizable Deepfake Detection

    Chuangchuang Tan, Renshuai Tao, Huan Liu, and Yao Zhao. GANGen-Detection: A Dataset Generated by GANs for Generalizable Deepfake Detection. github.com/chuangchuangtan/ GANGen-Detection, 2024

  29. [37]

    Frequency-Aware Deepfake Detection: Improving Generalizability through Frequency Space Domain Learning

    Chuangchuang Tan, Yao Zhao, Shikui Wei, Guanghua Gu, Ping Liu, and Yunchao Wei. Frequency-Aware Deepfake Detection: Improving Generalizability through Frequency Space Domain Learning. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 38, pages 5052–5060, 2024

  30. [38]

    Re- thinking the Up-Sampling Operations in CNN-Based Generative Network for Generalizable Deepfake Detection

    Chuangchuang Tan, Yao Zhao, Shikui Wei, Guanghua Gu, Ping Liu, and Yunchao Wei. Re- thinking the Up-Sampling Operations in CNN-Based Generative Network for Generalizable Deepfake Detection. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, p...

  31. [39]

    Learning on Gradients: Generalized Artifacts Representation for GAN-Generated Images Detection

    Chuangchuang Tan, Yao Zhao, Shikui Wei, Guanghua Gu, and Yunchao Wei. Learning on Gradients: Generalized Artifacts Representation for GAN-Generated Images Detection. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 12105–12114, 2023

  32. [40]

    Visualizing data using t-SNE

    Laurens Van der Maaten and Geoffrey Hinton. Visualizing data using t-SNE. Journal of Machine Learning Research, 9(11), 2008

  33. [41]

    Cnn- generated images are surprisingly easy to spot

    Sheng-Yu Wang, Oliver Wang, Richard Zhang, Andrew Owens, and Alexei A Efros. Cnn- generated images are surprisingly easy to spot... for now. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 8695–8704, 2020

  34. [42]

    DIRE for Diffusion-Generated Image Detection

    Zhendong Wang, Jianmin Bao, Wengang Zhou, Weilun Wang, Hezhen Hu, Hong Chen, and Houqiang Li. DIRE for Diffusion-Generated Image Detection. In Proceedings of the IEEE/CVF International Conference on Computer Vision, pages 22445–22455, 2023

  35. [43]

    Combating Misinformation in the Era of Generative AI Models

    Danni Xu, Shaojing Fan, and Mohan Kankanhalli. Combating Misinformation in the Era of Generative AI Models. InProceedings of the 31st ACM International Conference on Multimedia, MM ’23, page 9291–9298, 2023

  36. [44]

    LSUN: Construction of a Large-Scale Image Dataset using Deep Learning with Humans in the Loop

    Fisher Yu, Ari Seff, Yinda Zhang, Shuran Song, Thomas Funkhouser, and Jianxiong Xiao. LSUN: Construction of a Large-Scale Image Dataset using Deep Learning with Humans in the Loop. arXiv preprint arXiv:1506.03365, 2015

  37. [45]

    Breaking Semantic Artifacts for Generalized AI-generated Image Detection

    Chende Zheng, Chenhao Lin, Zhengyu Zhao, Hang Wang, Xu Guo, Shuai Liu, and Chao Shen. Breaking Semantic Artifacts for Generalized AI-generated Image Detection. In The Thirty-eighth Annual Conference on Neural Information Processing Systems, 2024

  38. [46]

    Patchcraft: Exploring Texture Patch for Efficient AI-generated Image Detection

    Nan Zhong, Yiran Xu, Sheng Li, Zhenxing Qian, and Xinpeng Zhang. Patchcraft: Exploring Texture Patch for Efficient AI-generated Image Detection. arXiv preprint arXiv:2311.12397, pages 1–18, 2024

  39. [47]

    Unpaired Image-to-Image Trans- lation using Cycle-Consistent Adversarial Networks

    Jun-Yan Zhu, Taesung Park, Phillip Isola, and Alexei A Efros. Unpaired Image-to-Image Trans- lation using Cycle-Consistent Adversarial Networks. In Proceedings of the IEEE international conference on computer vision, pages 2223–2232, 2017. 12

Pith tools

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