Pith. sign in

REVIEW 4 major objections 5 minor 1 cited by

Time Step Generating: A Universal Synthesized Deepfake Image Detector

T0 review · 4 major / 5 minor · reviewed 2026-08-12 · deepseek-v4-flash

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

Pith's one-line read A fixed, pretrained diffusion U-Net, evaluated on an image at a small timestep, produces noise-prediction features that let a ResNet-50 classify it as real or synthetic with 94.9% average cross-generator accuracy on GenImage.

desk verdict TSG is a fast, clever deepfake detector that works by classifying predicted noise from a frozen diffusion U-Net; the catch is that its 'universal' label rests on an ImageNet-to-ImageNet shortcut no one tests. read the letter →

arxiv 2411.11016 v2 pith:MG5TKPLJ submitted 2024-11-17 cs.CV cs.AI

Ziyue Zeng , Haoyuan Liu , Dingjie Peng , Luoxu Jing , Hiroshi Watanabe This is my paper · ORCID
classification cs.CVcs.AI
keywords deepfakedetectionsyntheticimagediffusionmodelnoisepredictionfeatureextractiontimestepGenResNet-50
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

The paper proposes Time Step Generating (TSG), a synthetic-image detector that feeds an image through a frozen U-Net from a pretrained diffusion model at a fixed timestep near zero and uses the network's noise prediction as a feature map for a ResNet-50 classifier. The authors claim this single forward pass exposes fine-grained differences between real and generated images without needing the slow inversion-and-reconstruction pipeline of prior methods. On five subsets of the GenImage benchmark the method averages 94.9% accuracy across generators, about 19 points above LaRE2, and runs roughly ten times faster than DIRE. If correct, TSG offers a fast, generator-agnostic way to detect diffusion-generated images.

What carries the argument

The central object is the TSG feature $F = \epsilon_\theta(I, t)$: the noise prediction of a frozen, class-unconditional ImageNet diffusion U-Net evaluated on the query image $I$ at a fixed timestep $t$ (typically 0). This is a single forward pass, in contrast to reconstruction-based detectors that invert and denoise. The timestep acts as a detail-preservation dial: at $t=0$ the predicted noise retains the finest high-frequency details, which the classifier uses; at $t=50$ some detail is lost and accuracy drops slightly. The paper interprets the mechanism through score-based diffusion: real images sit at high probability density, so the estimated score at their location differs from the score of generated images.

What would settle it

Take a set of real photographs that are clearly outside ImageNet's domain (e.g., medical scans or satellite imagery), run them through TSG with a classifier trained on ImageNet real images, and measure accuracy; if it falls to near chance while generated images from the same new domain are still detected, the universality claim fails. Alternatively, rerun the GenImage cross-validation after replacing the real half of the test set with non-ImageNet photographs and check whether the 94.9% average drops.

Watch

Extended reading notes

Core claim

On the paper's own terms, the central discovery is that the noise estimated by a pretrained diffusion U-Net at a small timestep t encodes a reliable real-versus-synthetic signal. Because real images already lie at high-density regions of the data distribution while generated images sit nearby but off the true manifold, the score (equivalently the predicted noise) at the same timestep differs systematically. TSG defines the feature as $F = \epsilon_\theta(I, t)$ and feeds $F$ into ResNet-50 for binary classification; no inversion, no denoising loop, and no task-specific generation model is required. The authors report 94.9% average accuracy across five generators and robustness to JPEG compression.

Load-bearing premise

The method is called universal but the frozen feature extractor was trained on ImageNet and every real image used in the experiments also comes from ImageNet, so the reported accuracy may rely on that distributional overlap and might drop on real photos from other sources.

Editorial extensions

If this is right

  • A single fixed extractor can serve multiple generators: classifiers trained on one diffusion model's output transfer to other diffusion models and to BigGAN, so a detector could be deployed without per-generator retraining.
  • Detection becomes cheap enough for screening: with about a tenth of DIRE's compute, one forward pass through a U-Net plus a ResNet-50 gives a label, making large-scale moderation more practical.
  • JPEG compression does not break the feature: cross-validation on compressed subsets of Glide, SD V1.4, and Midjourney stays high, so the method is applicable to images as shared online.
  • Training one classifier on a mix of GAN and diffusion outputs yields near-perfect accuracy on both families and on unseen compressed sets, suggesting a route to a single universal detector.
  • The timestep t is a tunable knob: t=0 is best for accuracy, while larger t trades detail for slightly lower but still strong performance.

Reading between the lines

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

  • The paper's 'universal' claim is tested only against real images from ImageNet, the same distribution the extractor was trained on; an editorially suggested extension is to measure TSG on real photographs from a different domain (e.g., camera-phone or medical imagery), where the score separation may shrink.
  • Because TSG exposes the U-Net's internal estimate of noise, the same machinery could be turned toward attribution—matching the noise pattern to the generator's fingerprint—rather than just binary real/fake.
  • The timestep sweep suggests a calibration strategy for other detectors: search over small t to maximize high-frequency detail, and use larger t to check robustness of a given classifier.
  • If the score-based interpretation is right, TSG should work with any sufficiently expressive diffusion backbone, not just ADM, which could be tested by swapping in Stable Diffusion or DiT U-Nets.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 5 minor

Summary. The paper proposes Time Step Generating (TSG), a synthetic image detector that uses a pretrained diffusion model U-Net as a fixed feature extractor. For an input image I and a chosen time step t, it computes the noise prediction F = ε_θ(I, t) and feeds F into a ResNet-50 binary classifier. The authors evaluate TSG on five GenImage subsets (BigGAN, VQDM, SD V1.5, ADM, Wukong), report an average cross-generator accuracy of 94.9% at t=0, and measure a roughly 10x speedup over DIRE. They also include JPEG-robustness experiments on three additional subsets and a mixed-training experiment achieving 100.0% accuracy on all test sets.

Significance. The core idea is simple and attractive: instead of running full diffusion inversion and reconstruction, one forward pass through a frozen diffusion U-Net at a small time step yields features that separate real from synthetic images. If the universality claim were supported, this would be a practical and fast detector with strong cross-generator generalization. The paper is also commendable for reporting code availability and for transparently comparing t=0 and t=50. However, the current evidence does not establish universality: the real test images are drawn from ImageNet, which is also the training distribution of the frozen feature extractor, and no experiment uses real photographs from a different distribution. The absence of error bars and the suspicious exactly-100.0% mixed-training result further limit the strength of the empirical claims.

major comments (4)
  1. [Sections 4.1 and 4.3] The feature extractor is the class-unconditional ImageNet diffusion model (ADM) released with [7], and the real images in all GenImage subsets are also ImageNet images. Consequently, in every experiment the real class is exactly the extractor's training distribution, while all synthetic images are out-of-distribution. The classifier may therefore learn to separate 'familiar ImageNet input' from 'unfamiliar OOD input' rather than a general real-vs-synthetic boundary. The abstract's claim that TSG 'does not rely on ... specific datasets' is directly contradicted by this setup, and the claimed universality is untested. Please add experiments with real photographs from a different distribution (e.g., COCO, LSUN, or a camera-captured dataset), or substantially temper the universality claim.
  2. [Tables 1 and 4, Sections 4.5 and 4.8] All accuracies reported in Table 1 and Table 4 come from a single train/validation split with no error bars or multiple seeds. More importantly, Table 4 reports exactly 100.0% accuracy on all test sets, including the Un-bias sets (Glide, SD V1.4, Midjourney), for a classifier trained on a mix of ADM, SD V1.5, and BigGAN. This perfect transfer is difficult to reconcile with the much lower cross-generator accuracies in Figure 4 (e.g., classifiers trained on BigGAN often perform poorly on diffusion subsets). The authors should explain the protocol, report mean and standard deviation over at least three seeds, and investigate whether a trivial cue (e.g., an averaging artifact in the feature extractor or a leak in the data pipeline) is responsible for the 100.0% numbers.
  3. [Section 4.6 and Table 3] The 'unbiased' datasets are constructed by filtering GenImage subsets (Glide, SD V1.4, Midjourney) by a JPEG 'compression rate greater than 96'. However, the real images in these subsets are still ImageNet images, so the robustness experiment only demonstrates invariance to JPEG compression within the same distribution that the feature extractor was trained on. The terminology 'unbiased' is therefore misleading, and the phrase 'compression rate greater than 96' is ambiguous (it could mean quality factor > 96 or file-size ratio). Please clarify the exact JPEG quality parameter and, if the universal-claim revision is pursued, test JPEG robustness on non-ImageNet real images as well.
  4. [Section 3.2 and Figure 2] The theoretical justification for why t=0 should be optimal is not fully developed. The score-based argument in Figure 2 posits that real images lie at high probability density and generated images at lower density, but this is not quantified and does not explain the empirical advantage of t=0 over t=50. The paper only compares two values of t and then fixes t=0; a systematic sweep over t (e.g., t=0, 10, 50, 100, 200) would strengthen the claim that the time step is a controllable and meaningful parameter rather than a selected hyperparameter.
minor comments (5)
  1. [Abstract and Section 1] There are typos: 'vary challenging' should be 'very challenging', and in the Introduction 'TGS simplifies the approach further' should be 'TSG'.
  2. [Table 2] The table formatting is broken; the column headers are unclear. Please present the DIRE and TSG rows with clear labels for the number of sampling steps, batch size, and measured time.
  3. [Section 4.6] The phrase 'compression rate greater than 96' should be specified precisely, e.g., 'JPEG quality factor > 96' or 'compression ratio > 96%', since it is currently ambiguous.
  4. [Section 4.4] The discussion of Figure 4 would benefit from explicit numbers for the t=50 average accuracy, since Table 1 reports only the aggregated values for t=0 and t=50.
  5. [Conclusion] The conclusion states the method is '19% better than LaRE2', but the table reports 94.9 vs. 75.6, which is 19.3 percentage points. Please use 'percentage points' to avoid confusion.

Circularity Check

0 steps flagged · score 1.0 of 10

No significant circularity: TSG is an empirical feature-extractor-plus-classifier pipeline; the central claim does not reduce to its inputs by construction.

full rationale

The derivation chain is self-contained against the circularity patterns. TSG defines F = eps_theta(I, t) (Eq. 8) using a frozen pretrained U-Net and then trains a ResNet-50 on labeled real/fake features; the reported accuracy is an empirical outcome, not a restatement of Eq. 8 or of any fitted parameter. The only tuned quantity, t, is compared transparently at two values (t = 0 and t = 50) and both are reported, so this is model selection rather than a fitted input disguised as a prediction. Reference [26] is a self-citation by a co-author, but it supports only the background statement that face-forgery detection work exists and is not load-bearing. The paper's "universal" claim is under-supported because, as stated in Section 4.1, "we adopt the class-unconditional ImageNet diffusion model," and no non-ImageNet real-photo test is reported; the real class may therefore be in-distribution for the extractor. That is an external-validity and correctness risk, not a circular derivation.

Assumptions & free parameters 2 free parameters · 4 assumptions · 0 invented entities

The method introduces no new physical or mathematical entities. It depends on the frozen U-Net being a suitable feature extractor for real images, on the informal score-density assumption, and on two hand-picked thresholds (t and JPEG quality). The central empirical claim is plausible but rests on evaluation choices that are not fully justified.

free parameters (2)
  • time step t = 0 (headline), 50 (secondary)
    The paper evaluates t=0 and t=50 and uses t=0 for the headline results in Table 1. The choice is justified only by the empirical comparison in Section 4.4, not by a theoretical derivation.
  • JPEG quality threshold = 96 (greater than 96 selected)
    In Section 4.6, the robustness test sets are built by selecting GenImage images with JPEG compression rate greater than 96. The cutoff is arbitrary and not motivated by any stated principle.
assumptions (4)
  • domain assumption The frozen ADM U-Net trained on ImageNet is an approximately correct score estimator for the image distribution.
    Section 4.1 adopts this network as the fixed feature extractor; Section 3.2 relies on score-based reasoning about real versus generated samples.
  • domain assumption Real images sit at high-density points of the score model's distribution and generated images do not, making their noise predictions distinguishable.
    This is the informal motivation in Section 3.2; it is asserted rather than verified with any direct measurement.
  • domain assumption The GenImage real set, which is ImageNet, is treated as representative of all real images.
    All experiments use GenImage's split, whose real images are ImageNet photos (Section 4.1). The paper never tests on real photos outside ImageNet, so the universal claim rests on this representativeness assumption.
  • ad hoc to paper A predicted noise map at t=0 or t=50 carries enough information for real-versus-fake binary classification.
    This is the definitional premise of the method (Equation 8, Section 3.2); its validity is demonstrated only empirically on the selected subsets.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Time Step Generating: A Universal Synthesized Deepfake Image Detector." pith.science (2026). https://pith.science/paper/MG5TKPLJ

@misc{pith2026241111016,
  author       = {Pith},
  title        = {Pith review of: Time Step Generating: A Universal Synthesized Deepfake Image Detector},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/MG5TKPLJ}},
  note         = {Machine review of arXiv:2411.11016}
}
read the original abstract

Currently, high-fidelity text-to-image models are developed in an accelerating pace. Among them, Diffusion Models have led to a remarkable improvement in the quality of image generation, making it vary challenging to distinguish between real and synthesized images. It simultaneously raises serious concerns regarding privacy and security. Some methods are proposed to distinguish the diffusion model generated images through reconstructing. However, the inversion and denoising processes are time-consuming and heavily reliant on the pre-trained generative model. Consequently, if the pre-trained generative model meet the problem of out-of-domain, the detection performance declines. To address this issue, we propose a universal synthetic image detector Time Step Generating (TSG), which does not rely on pre-trained models' reconstructing ability, specific datasets, or sampling algorithms. Our method utilizes a pre-trained diffusion model's network as a feature extractor to capture fine-grained details, focusing on the subtle differences between real and synthetic images. By controlling the time step t of the network input, we can effectively extract these distinguishing detail features. Then, those features can be passed through a classifier (i.e. Resnet), which efficiently detects whether an image is synthetic or real. We test the proposed TSG on the large-scale GenImage benchmark and it achieves significant improvements in both accuracy and generalizability.

Figures

Figures reproduced from arXiv: 2411.11016 by the authors.

Figure 1
Figure 1. Overview of the reconstructing based method and our [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 3
Figure 3. The feature images output by TSG under different con [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figure 4
Figure 4. Cross validation results on various training and testing subsets of Genimage. For each generator, a model is trained and tested [PITH_FULL_IMAGE:figures/full_fig_p005_4.png] view at source ↗
Figures from the paper (3 more)
Figure 5
Figure 5. Figure 5: Using Grad-CAM heatmaps to demonstrate the part classifier relies for classification. [PITH_FULL_IMAGE:figures/full_fig_p006_5.png]
Figure 6
Figure 6. Figure 6: Cross-validation on an unbiased datasets. [PITH_FULL_IMAGE:figures/full_fig_p006_6.png]
Figure 7
Figure 7. Figure 7: Grad-CAM visualizations for the same image at [PITH_FULL_IMAGE:figures/full_fig_p007_7.png]

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Towards Reliable Identification of Diffusion-based Image Manipulations

    cs.CV 2025-06 conditional novelty 6.0 of 10

    RADAR combines semantic and geometric vision features with contrastive learning to detect and localize diffusion-based image edits, outperforming prior methods on a new 28-model benchmark.

Reference graph

Works this paper leans on

42 extracted references · 23 canonical work pages · cited by 1 Pith paper

  1. [7]

    Diffusion models beat gans on image synthesis

    Prafulla Dhariwal and Alexander Nichol. Diffusion models beat gans on image synthesis. In M. Ranzato, A. Beygelz- imer, Y . Dauphin, P.S. Liang, and J. Wortman Vaughan, ed- itors, Advances in Neural Information Processing Systems , volume 34, pages 8780–8794. Curran Associates, Inc., 2021. 1, 4

    work page 2021

  2. [1]

    Denoising dif- fusion probabilistic models

    Jonathan Ho, Ajay Jain, and Pieter Abbeel. Denoising dif- fusion probabilistic models. Advances in neural information processing systems, 33:6840–6851, 2020. 1, 2

    2020

  3. [2]

    Denois- ing diffusion implicit models

    Jiaming Song, Chenlin Meng, and Stefano Ermon. Denois- ing diffusion implicit models. In International Conference on Learning Representations, 2021. 1

    2021

  4. [3]

    Towards more accurate diffusion model acceleration with a timestep tuner

    Mengfei Xia, Yujun Shen, Changsong Lei, Yu Zhou, Deli Zhao, Ran Yi, Wenping Wang, and Yong-Jin Liu. Towards more accurate diffusion model acceleration with a timestep tuner. In Proceedings of the IEEE/CVF Conference on Com- puter Vision and Pattern Recognition , pages 5736–5745,

    work page

  5. [4]

    Attention-driven training-free efficiency enhancement of diffusion models

    Hongjie Wang, Difan Liu, Yan Kang, Yijun Li, Zhe Lin, Ni- raj K Jha, and Yuchen Liu. Attention-driven training-free efficiency enhancement of diffusion models. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pat- tern Recognition, pages 16080–16089, 2024. 1

    work page 2024

  6. [5]

    Dif- fusion models without attention

    Jing Nathan Yan, Jiatao Gu, and Alexander M Rush. Dif- fusion models without attention. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 8239–8249, 2024. 1

    work page 2024

  7. [6]

    Improving train- ing efficiency of diffusion models via multi-stage framework and tailored multi-decoder architecture

    Huijie Zhang, Yifu Lu, Ismail Alkhouri, Saiprasad Ravis- hankar, Dogyoon Song, and Qing Qu. Improving train- ing efficiency of diffusion models via multi-stage framework and tailored multi-decoder architecture. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 7372–7381, 2024. 1

    2024

  8. [8]

    Deep unsupervised learning using nonequilibrium thermodynamics

    Jascha Sohl-Dickstein, Eric Weiss, Niru Maheswaranathan, and Surya Ganguli. Deep unsupervised learning using nonequilibrium thermodynamics. In International confer- 7 ence on machine learning, pages 2256–2265. PMLR, 2015. 1

    work page 2015

Show all 42 references
  1. [9]

    High-resolution image synthesis with latent diffusion models

    Robin Rombach, Andreas Blattmann, Dominik Lorenz, Patrick Esser, and Bj ¨orn Ommer. High-resolution image synthesis with latent diffusion models. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pages 10684–10695, June 2022. 1, 2, 4

    2022

  2. [10]

    Improved denoising diffusion probabilistic models

    Alexander Quinn Nichol and Prafulla Dhariwal. Improved denoising diffusion probabilistic models. In International conference on machine learning, pages 8162–8171. PMLR,

  3. [11]

    Classifier-free diffusion guidance

    Jonathan Ho and Tim Salimans. Classifier-free diffusion guidance. arXiv preprint arXiv:2207.12598, 2022. 1

    2022 arXiv

  4. [12]

    Video dif- fusion models

    Jonathan Ho, Tim Salimans, Alexey Gritsenko, William Chan, Mohammad Norouzi, and David J Fleet. Video dif- fusion models. Advances in Neural Information Processing Systems, 35:8633–8646, 2022. 1

    2022

  5. [13]

    Controlnet ++: Improving conditional controls with efficient consistency feedback

    Ming Li, Taojiannan Yang, Huafeng Kuang, Jie Wu, Zhaon- ing Wang, Xuefeng Xiao, and Chen Chen. Controlnet ++: Improving conditional controls with efficient consistency feedback. In European Conference on Computer Vision , pages 129–147. Springer, 2025. 1

    2025

  6. [14]

    Controlnext: Powerful and effi- cient control for image and video generation

    Bohao Peng, Jian Wang, Yuechen Zhang, Wenbo Li, Ming- Chang Yang, and Jiaya Jia. Controlnext: Powerful and effi- cient control for image and video generation. arXiv preprint arXiv:2408.06070, 2024. 1

    2024 arXiv

  7. [15]

    Imagic: Text-based real image editing with diffusion models

    Bahjat Kawar, Shiran Zada, Oran Lang, Omer Tov, Huiwen Chang, Tali Dekel, Inbar Mosseri, and Michal Irani. Imagic: Text-based real image editing with diffusion models. InPro- ceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 6007–6017, 2023. 1

    2023

  8. [16]

    In- structpix2pix: Learning to follow image editing instructions

    Tim Brooks, Aleksander Holynski, and Alexei A Efros. In- structpix2pix: Learning to follow image editing instructions. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 18392–18402, 2023. 1

    2023

  9. [17]

    Countering malicious deepfakes: Survey, battleground, and horizon

    Felix Juefei-Xu, Run Wang, Yihao Huang, Qing Guo, Lei Ma, and Yang Liu. Countering malicious deepfakes: Survey, battleground, and horizon. International journal of computer vision, 130(7):1678–1734, 2022. 1

    2022

  10. [18]

    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 (ICCV), pages 22445–22455, October 2023. 2, 3, 5

    2023

  11. [19]

    Lareˆ2: Latent reconstruction error based method for diffusion-generated image detection

    Yunpeng Luo, Junlong Du, Ke Yan, and Shouhong Ding. Lareˆ2: Latent reconstruction error based method for diffusion-generated image detection. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pages 17006–17015, June 2024. 2, 3, 5

    2024

  12. [20]

    Genimage: A million-scale benchmark for de- tecting ai-generated image

    Mingjian Zhu, Hanting Chen, Qiangyu Yan, Xudong Huang, Guanyu Lin, Wei Li, Zhijun Tu, Hailin Hu, Jie Hu, and Yunhe Wang. Genimage: A million-scale benchmark for de- tecting ai-generated image. Advances in Neural Information Processing Systems, 36, 2024. 2, 4

    2024

  13. [21]

    Generative adversarial nets

    Ian Goodfellow, Jean Pouget-Abadie, Mehdi Mirza, Bing Xu, David Warde-Farley, Sherjil Ozair, Aaron Courville, and Yoshua Bengio. Generative adversarial nets. In Z. Ghahra- mani, M. Welling, C. Cortes, N. Lawrence, and K.Q. Wein- berger, editors, Advances in Neural Information ...

    2014

  14. [22]

    Auto-encoding variational bayes

    Diederik P Kingma. Auto-encoding variational bayes. arXiv preprint arXiv:1312.6114, 2013. 2

    2013 arXiv

  15. [23]

    Pixart-$\alpha$: Fast training of diffusion transformer for photorealistic text-to-image syn- thesis

    Junsong Chen, Jincheng YU, Chongjian GE, Lewei Yao, Enze Xie, Zhongdao Wang, James Kwok, Ping Luo, Huchuan Lu, and Zhenguo Li. Pixart-$\alpha$: Fast training of diffusion transformer for photorealistic text-to-image syn- thesis. In The Twelfth International Conference on Learn...

    2024

  16. [24]

    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. 3

    2020

  17. [25]

    What makes fake images detectable? understanding proper- ties that generalize

    Lucy Chai, David Bau, Ser-Nam Lim, and Phillip Isola. What makes fake images detectable? understanding proper- ties that generalize. In Computer Vision–ECCV 2020: 16th European Conference, Glasgow, UK, August 23–28, 2020, Proceedings, Part XXVI 16, pages 103–120. Springer, 2020. 3

    2020

  18. [26]

    Adversarial level of face images generated by prompt-based image coding in face recognition system

    Yurika Fujinami Hiroshi Watanabe. Adversarial level of face images generated by prompt-based image coding in face recognition system. In IEEE Global Conference on Con- sumer Electronics (GCCE2024)., pages 332–333, 2024. 3

    2024

  19. [27]

    Poisoned forgery face: Towards backdoor attacks on face forgery detection

    Jiawei Liang, Siyuan Liang, Aishan Liu, Xiaojun Jia, Junhao Kuang, and Xiaochun Cao. Poisoned forgery face: Towards backdoor attacks on face forgery detection. arXiv preprint arXiv:2402.11473, 2024. 3

    2024 arXiv

  20. [28]

    Improving gan-generated image detec- tion generalization using unsupervised domain adaptation

    Mingxu Zhang, Hongxia Wang, Peisong He, Asad Malik, and Hanqing Liu. Improving gan-generated image detec- tion generalization using unsupervised domain adaptation. In 2022 IEEE International Conference on Multimedia and Expo (ICME), pages 1–6. IEEE, 2022. 3

    2022

  21. [29]

    Whodunit: Detection and attribution of synthetic images by leveraging model-specific fingerprints

    Alexander Wißmann, Steffen Zeiler, Robert M Nickel, and Dorothea Kolossa. Whodunit: Detection and attribution of synthetic images by leveraging model-specific fingerprints. In Proceedings of the 3rd ACM International Workshop on Multimedia AI against Disinformation , pages 65–...

    2024

  22. [30]

    Rigid: A training-free and model-agnostic framework for ro- bust ai-generated image detection

    Zhiyuan He, Pin-Yu Chen, and Tsung-Yi Ho. Rigid: A training-free and model-agnostic framework for ro- bust ai-generated image detection. arXiv preprint arXiv:2405.20112, 2024. 3

    2024 arXiv

  23. [31]

    Generalizable synthetic image detection via language-guided contrastive learning

    Haiwei Wu, Jiantao Zhou, and Shile Zhang. Generalizable synthetic image detection via language-guided contrastive learning. arXiv preprint arXiv:2305.13800, 2023. 3

    2023 arXiv

  24. [32]

    Learning transferable visual models from natural language supervi- sion

    Alec Radford, Jong Wook Kim, Chris Hallacy, Aditya Ramesh, Gabriel Goh, Sandhini Agarwal, Girish Sastry, Amanda Askell, Pamela Mishkin, Jack Clark, et al. Learning transferable visual models from natural language supervi- sion. In International conference on machine learning, ...

    2021

  25. [33]

    Fakeinversion: Learning to detect images from un- seen text-to-image models by inverting stable diffusion

    George Cazenavette, Avneesh Sud, Thomas Leung, and Ben Usman. Fakeinversion: Learning to detect images from un- seen text-to-image models by inverting stable diffusion. In Proceedings of the IEEE/CVF Conference on Computer Vi- sion and Pattern Recognition (CVPR), pages 10759–1...

    2024

  26. [34]

    Learning on gradients: Generalized arti- facts representation for gan-generated images detection

    Chuangchuang Tan, Yao Zhao, Shikui Wei, Guanghua Gu, and Yunchao Wei. Learning on gradients: Generalized arti- facts representation for gan-generated images detection. In Proceedings of the IEEE/CVF Conference on Computer Vi- sion and Pattern Recognition (CVPR), pages 12105–12...

    2023

  27. [35]

    Deep residual learning for image recognition

    Kaiming He, Xiangyu Zhang, Shaoqing Ren, and Jian Sun. Deep residual learning for image recognition. In Proceed- ings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), June 2016. 4

    2016

  28. [36]

    Large scale GAN training for high fidelity natural image synthe- sis

    Andrew Brock, Jeff Donahue, and Karen Simonyan. Large scale GAN training for high fidelity natural image synthe- sis. In International Conference on Learning Representa- tions, 2019. 4

    2019

  29. [37]

    Vec- tor 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. Vec- tor quantized diffusion model for text-to-image synthesis. In Proceedings of the IEEE/CVF conference on computer vi- sion and pattern recognition, pages 10696–10706, 2022. 4

    2022

  30. [38]

    Wukong: A 100 million large-scale chinese cross-modal pre-training benchmark

    Jiaxi Gu, Xiaojun Meng, Guansong Lu, Lu Hou, Niu Minzhe, Xiaodan Liang, Lewei Yao, Runhui Huang, Wei Zhang, Xin Jiang, et al. Wukong: A 100 million large-scale chinese cross-modal pre-training benchmark. Advances in Neural Information Processing Systems , 35:26418–26431,

  31. [39]

    Zhengzhe Liu, Xiaojuan Qi, and Philip H.S. Torr. Global texture enhancement for fake face detection in the wild. In Proceedings of the IEEE/CVF Conference on Computer Vi- sion and Pattern Recognition (CVPR), June 2020. 5

    2020

  32. [40]

    Fake or jpeg? revealing common bi- ases in generated image detection datasets

    Patrick Grommelt, Louis Weiss, Franz-Josef Pfreundt, and Janis Keuper. Fake or jpeg? revealing common bi- ases in generated image detection datasets. arXiv preprint arXiv:2403.17608, 2024. 6

    2024 arXiv

  33. [41]

    Glide: Towards photorealistic image generation and editing with text-guided diffusion models.arXiv preprint arXiv:2112.10741, 2021

    Alex Nichol, Prafulla Dhariwal, Aditya Ramesh, Pranav Shyam, Pamela Mishkin, Bob McGrew, Ilya Sutskever, and Mark Chen. Glide: Towards photorealistic image generation and editing with text-guided diffusion models.arXiv preprint arXiv:2112.10741, 2021. 6

    2021 arXiv

  34. [42]

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

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

Pith tools

Reviewed August 12, 2026 · model on record in the stance chip above.