Pith. sign in

REVIEW 4 major objections 6 minor 30 references

CTForensics: A Comprehensive Dataset and Method for AI-Generated CT Image Detection

T0 review · 4 major / 6 minor · reviewed 2026-08-02 · deepseek-v4-flash

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

Pith's one-line read This paper claims that a detector built around CT-specific wavelet, spatial, and frequency cues can identify AI-generated CT images with 96% accuracy across ten generative models, outperforming prior methods by more than 10 percentage point

desk verdict Useful new CT forgery benchmark, but the headline generalization claim is undercut by a domain confound and a train/test generator overlap. read the letter →

arxiv 2603.01878 v2 pith:SFBVOANN submitted 2026-03-02 cs.CV

Yiheng Li , Zichang Tan , Guoqing Xu , Yichun Yeh , Yang Yang , Zhen Lei This is my paper · ORCID
classification cs.CV
keywords AI-generatedimagedetectionCTforensicsmedicaldeepfakescomputedtomographygenerativeadversarialnetworksdiffusionmodelswavelettransformfrequencyanalysis
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 sets out to show that AI-generated CT images can be reliably detected when the detector is built around CT-specific imaging physics rather than adapted from natural-photo forensics. To test this, it introduces CTForensics, a benchmark of 75,990 2D CT slices with a balanced test set of 29,990 images from ten generative models spanning GANs and diffusion models. It then presents ESF-CTFD, a convolutional neural network that combines wavelet, spatial, and frequency analysis across three image scales. The paper reports 96.01% mean accuracy and 99.96% mean average precision on this benchmark, with only a 0.99% average drop under blur, cropping, JPEG compression, and noise. If correct, this gives clinical settings a practical way to flag synthetic scans and a shared benchmark to compare future detectors.

What carries the argument

The key mechanism is a three-branch architecture tuned to CT-specific traces: a Wavelet-Enhanced Central Stem decomposes the image into sub-bands and amplifies local intensity differences; Multi-Scale Spatial Aggregation fuses features at 112, 224, and 448 pixel resolutions; and a Frequency Process Block applies Fast Fourier Convolution to expose spectral artifacts. The CTForensics benchmark itself is the enabling asset, providing ten generative models (four GAN-based, six diffusion-based) with balanced real-fake test sets, so generalisation can actually be measured.

What would settle it

Run the same training/evaluation protocol but replace the real test images with CT slices from a different institution or scanner (keeping the same ten generators); if mean accuracy drops sharply while the detector still tags the training generator perfectly, the model was keyed to dataset-specific cues, not universal forgery artifacts. A patient-level split of CTForensics would also show whether overlapping patients inflate the scores.

Watch

Extended reading notes

Core claim

The central claim is that a detector explicitly modeling CT artifacts in three complementary domains—high-frequency wavelet residuals, local pixel correlations, and global spectral statistics—generalizes across unseen generative models. Trained on fake images from a single GAN, ESF-CTFD achieves 96.01% mean accuracy and 99.96% mean average precision on the CTForensics test set, which includes nine additional generators, and outperforms existing natural-image forensics methods by more than 10 percentage points. The paper further claims the design makes the detector resilient to realistic distortions, with only a 0.99% average accuracy drop under Gaussian blur, random cropping, JPEG compressio

Load-bearing premise

The load-bearing premise is that 'real' CT images are well represented by a single public dataset, so the detector can learn a general real-image signature; if that dataset's quirks—noise, windowing, anatomy—are what the model exploits, the reported accuracy will not carry over to other hospitals or scanners.

Editorial extensions

If this is right

  • A model trained on fake images from a single GAN can identify images from nine unseen generators, covering both GAN and diffusion architectures.
  • Combining wavelet, spatial, and frequency cues yields large gains over any single branch: removing both wavelet and central-correlation components drops accuracy from 96.01% to 88.37%.
  • Multi-scale processing (112, 224, and 448 pixels) is necessary; dropping any scale reduces both mean accuracy and mean average precision.
  • The detector loses only 0.99% accuracy on average under Gaussian blur, random cropping, JPEG compression, and injected noise, even when all are applied together.
  • CTForensics provides a fixed benchmark with ten generative models, enabling direct comparison of future CT forgery detectors.

Reading between the lines

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

  • Because all real images come from a single public CT dataset, the 96% figure may partly encode a dataset fingerprint; testing on real scans from other scanners or hospitals would show whether the detector learns a general 'real' distribution.
  • The training generator also appears in the test set, so the perfect 100% score on it may lift the mean; a leave-one-generator-out protocol would isolate true cross-generator generalization.
  • The frequency-artifact analysis is tied to the ten generators studied; new generators with different spectral signatures, or post-processing that alters the spectrum, could require the benchmark to be refreshed.
  • The same wavelet–spatial–frequency architecture could transfer to other modalities where synthetic images are a growing concern, such as MRI or X-ray, if real data are collected from multiple sources.
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 / 6 minor

Summary. The paper introduces CTForensics, a benchmark of 75,990 2D CT images with a balanced 29,990-image test set spanning ten generative models (four GAN-based, six diffusion-based). It also proposes ESF-CTFD, a CNN combining a wavelet-enhanced stem, multi-scale spatial aggregation, and a frequency-process block. On the authors' benchmark, ESF-CTFD reports 96.01% mean accuracy and 99.96% mean average precision, outperforming existing detectors by a large margin, and shows only a 0.99% average mAcc drop under blur, cropping, JPEG compression, and noise. The paper's central claim is that ESF-CTFD generalizes across unseen CT generative models and is robust to realistic perturbations.

Significance. A multi-generator CT forgery benchmark and a CT-specific detector would be a useful contribution if the evaluation protocol substantiated the generalization claim. The paper offers per-generator results, ablations of the multi-scale and wavelet/central components, and a robustness comparison. However, the current protocol does not yet support the headline claim: the training fake-generator HA-GAN appears among the test generators, all real images are drawn from a single source (CT-RATE) while fakes come from heterogeneous datasets, and no patient-level split is reported. The dataset has potential, and the method's internal design is coherent, but the evidence for unseen-generator and cross-domain generalization needs to be established by a revised evaluation.

major comments (4)
  1. [Sec. 3 vs. Sec. 4 and Table 1] The task definition in Sec. 3 states that the model is trained on fake images from one generative method and evaluated on multiple unseen forgery types. Sec. 4 says the training fakes are generated using HA-GAN, yet HA-GAN is also one of the ten test generators in Table 1, where it achieves 100.00% accuracy. This makes the 'unseen generator' evaluation circular for one test column. Remove HA-GAN from the test set, retrain on a different generator, or report mAcc excluding the training generator; otherwise the 96.01% figure is not an unbiased generalization estimate.
  2. [Sec. 4, Table 1, and Fig. 1] All real images in both training and testing come from CT-RATE, while the fake images from the ten generators were produced by models trained on other CT datasets (e.g., LIDC, COVID-CT). The real-vs-fake decision is therefore confounded with source domain: a detector can succeed by recognizing CT-RATE-specific intensity statistics, windowing, or anatomy rather than generic generative artifacts. To support the claimed generalization, evaluate the same detector on real images from a second dataset (and, ideally, pair each generator with the held-out real domain), or explicitly control for source domain.
  3. [Sec. 4 (train/test split)] No patient-level split is reported. The 23,000 real training slices and the real test slices are both from CT-RATE, and if slices from the same patients appear in both sets, the reported accuracy is inflated by patient-level leakage. Please specify the patient/volume-level split and verify that no patient overlaps between training and test; if such a split already exists, state it explicitly.
  4. [Sec. 5.4 and Fig. 3] The frequency-process block is motivated by average frequency spectra computed over the test images shown in Fig. 3. Using test-set statistics to select the architecture or feature transformations means the reported generalization is not fully blind. This should be acknowledged as a limitation, and the method should be validated on a separate, untouched test set or an external dataset.
minor comments (6)
  1. [Sec. 1; Sec. 6.2] 'It comprises of there components' should be 'three components'; the Sec. 6.2 heading 'Performamce' should be 'Performance'.
  2. [Sec. 5.2] The text says 'hh, hl, and lh sub-bands ... while lh is handled by WTConv'; lh is listed twice and the low-frequency sub-band should be ll. Correct the notation.
  3. [Sec. 6.1] Define mAcc and mAP precisely. Clarify whether mAcc is the average of per-generator accuracies and whether the real test images are shared across generators.
  4. [Table 5] State whether perturbations are applied to both real and fake images, and give the exact ranges in the caption. Also 'A verage Drop' contains a typo.
  5. [Sec. 4 and Sec. 6.1] Report the CT preprocessing pipeline (windowing, intensity normalization, PNG conversion) and any filtering/quality checks used in dataset construction; this is needed for reproducibility.
  6. [Tables 1, 2, 5] All tables report point estimates without confidence intervals. Consider repeated runs with mean and standard deviation, especially for the per-generator results.

Circularity Check

2 steps flagged · score 4.0 of 10

Partial evaluation circularity: HA-GAN is used to generate training fakes and also appears in the test set, and the frequency block is motivated by spectra from the same benchmark, but the core method is not fitted to its own outputs.

  1. fitted input called prediction [Sec. 3 (Problem Definition); Sec. 4 (Dataset); Table 1]
    "we follow the widely adopted generalization paradigm in AI-generated image detection [20], where the model is trained using fake images from a single generative method, while being evaluated on multiple unseen forgery types at test time. ... Fake images are generated using HA-GAN [24] ... For the testing set, we employ ten generative models ... including ... HA-GAN [24] ... (Table 1: Ours ... HA-GAN ... 100.00)."

    The claimed 'unseen forgery types' evaluation includes HA-GAN, the very generator used to produce the training fakes. The 100.00% accuracy on HA-GAN in Table 1 is therefore an in-distribution result, not a cross-generator prediction. Averaging it into mAcc (96.01%) and mAP (99.96%) inflates the headline generalization numbers; the paper's own dataset construction makes this column trivially easy.

  2. other [Sec. 5.4 (Frequency Process Block), Fig. 3]
    "we visualize the frequency analysis maps of images generated by each generative model and real images, where the frequency map for each category is obtained by averaging the spectral information of multiple images in that category. It can be observed that AI-generated images exhibit significant artifacts in the frequency spectrum, which inspires us to further improve the model’s ability through frequency-domain feature processing."

    The FPB design is justified by averaged frequency spectra from the same benchmark categories on which the model is later evaluated. This is test-set-informed architecture selection, not a first-principles derivation: the frequency-processing choice is made after seeing the test set's spectral properties, so the subsequent comparison on that test set does not independently validate the frequency hypothesis.

full rationale

There is no equation-level circularity: Eq. (1)-(3) are standard constructions (wavelet convolution, central-difference expansion, FFC residual block), and the model is trained with binary entropy loss on a fixed split. However, the evaluation of the central generalization claim is partially circular. First, HA-GAN is used both to create training fakes and appears in the test set, contradicting the stated 'unseen forgery types' paradigm; the HA-GAN column is self-referential. Second, the frequency-processing design is motivated by averaged spectra from the same benchmark categories used for testing, so the frequency hypothesis is not independently tested. The domain confound (all real images from CT-RATE while fakes come from heterogeneous datasets) is a serious external-validity concern, but it is not a derivation circularity. Because the remaining nine generators are genuinely unseen and the ablations show internal consistency, the circularity is partial, not total.

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

The method introduces no new entities and has few free parameters. The central claims rest instead on dataset-composition assumptions — CT-RATE representing all authentic CTs, HA-GAN representing all generators, and no patient-level leakage — plus the legitimacy of using test-set statistics to design the architecture.

free parameters (2)
  • Input scale set = 112×112, 224×224, 448×448
    Chosen by ablation directly on the test benchmark (Table 3); the three-scale combination yields the best mAcc/mAP, so this is test-set-driven model selection.
  • Accuracy threshold = 0.5
    The threshold used to compute Acc in all tables; hand-set standard value with no operating-point or ROC analysis.
assumptions (4)
  • domain assumption AI-generated CT images leave detectable artifacts in wavelet, spatial, and frequency domains that can be learned from a single training generator (HA-GAN) and transferred to ten unseen generators.
    Underlies the entire generalization protocol (Sec. 3-4); no evidence is provided that HA-GAN's artifact space spans the diffusion models like StableDiffusion, GenerateCT, or Text2CT.
  • domain assumption CT-RATE images are an unbiased 'authentic' class for all ten fake generators.
    All real train and test images come from CT-RATE (Sec. 4), while many fake generators were trained on other CT distributions; differences in noise, resolution, and anatomy can be exploited instead of forgery cues.
  • domain assumption 2D slices extracted from 3D volumes are statistically independent, with no patient-level overlap between train and test.
    The paper reports random slice extraction (Sec. 4) without describing a patient-level split; if slices from the same 3D volume appear in both splits, memorization inflates the reported accuracy.
  • ad hoc to paper The average frequency spectrum of the test set (Fig. 3) is a legitimate design signal that will not bias generalization estimates.
    The FPB architecture is motivated by test-set frequency statistics; using test information to design the model invalidates the test set as an unbiased estimate of generalization.

how reviews work

0 comments
Cite this review

Pith. "Pith review of CTForensics: A Comprehensive Dataset and Method for AI-Generated CT Image Detection." pith.science (2026). https://pith.science/paper/SFBVOANN

@misc{pith2026260301878,
  author       = {Pith},
  title        = {Pith review of: CTForensics: A Comprehensive Dataset and Method for AI-Generated CT Image Detection},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/SFBVOANN}},
  note         = {Machine review of arXiv:2603.01878}
}
read the original abstract

Recent advances in generative AI have made synthetic Computed Tomography (CT) images increasingly realistic, enabling promising applications in medical data augmentation while raising serious concerns about clinical safety and data trustworthiness. Detecting AI-generated CT images remains challenging for two key reasons: existing benchmarks cover only limited generation sources, and many detectors are adapted from natural-image forensics without explicitly modeling CT-specific imaging properties. In this paper, we introduce CTForensics, a dataset for detecting AI-generated CT images. CTForensics contains 75,990 2D CT images, including a dedicated test benchmark of 29,990 balanced authentic and generated samples from ten representative CT generative models spanning GAN-based and diffusion-based paradigms. We further propose the Enhanced Spatial-Frequency CT Forgery Detector (ESF-CTFD), a CT-oriented CNN framework built around a Wavelet-Enhanced Central Stem, Multi-Scale Spatial Aggregation, and a Frequency-Aware Prediction Block. The Wavelet-Enhanced Central Stem enhances local intensity correlations and high-frequency residuals, Multi-Scale Spatial Aggregation aligns anatomical features across resolutions with lightweight residual units, and the Frequency-Aware Prediction Block models global spectral artifacts. Extensive experiments on CTForensics show that ESF-CTFD achieves 96.01% mAcc and 99.96% mAP, outperforming existing methods and maintaining strong robustness under realistic perturbations with only a 0.99% average drop. Codes will be available at https://github.com/liyih/CTForensics.

Figures

Figures reproduced from arXiv: 2603.01878 by the authors.

Figure 1
Figure 1. Visualization of CTForensics Dataset. 23000 2D CT slices from the original 3D CT releases. Fake images are generated using HA-GAN [24], with an equal number of real and synthetic images. For the testing set, we employ ten generative models for image generation, in￾cluding CGAN [17], CTGAN [18], sRD-GAN [12], GenerateCT [6], LungDDPM [9], LungDDPM+ [10], MedSyn [28], HA-GAN [24], StableDiffusion [22], and Text2CT [19… view at source ↗
Figure 2
Figure 2. Overall framework of our proposed ESF-CTFD. [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗
Figure 3
Figure 3. Frequency Analysis of CTForensics Dataset. [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

30 extracted references · 2 linked inside Pith

  1. [1]

    Multimedia Tools and Applications83(16), 48357–48375 (2024)

    Albahli, S., Nawaz, M.: Mednet: Medical deepfakes detection using an improved deep learning approach. Multimedia Tools and Applications83(16), 48357–48375 (2024)

    2024

  2. [2]

    In: Neural Information Pro- cessing Systems (2020)

    Chi, L., Jiang, B., Mu, Y.: Fast fourier convolution. In: Neural Information Pro- cessing Systems (2020)

    2020

  3. [3]

    In: European Conference on Computer Vision

    Finder, S.E., Amoyal, R., Treister, E., Freifeld, O.: Wavelet convolutions for large receptive fields. In: European Conference on Computer Vision. pp. 363–380. Springer (2024)

    2024

  4. [4]

    Advances in neural in- formation processing systems27(2014)

    Goodfellow, I.J., Pouget-Abadie, J., Mirza, M., Xu, B., Warde-Farley, D., Ozair, S., Courville, A., Bengio, Y.: Generative adversarial nets. Advances in neural in- formation processing systems27(2014)

    2014

  5. [5]

    ACM Transactions on Intelligent Systems and Technology16(6), 1–26 (2025)

    Grabovski, F.M., Yasur, L., Amit, G., Mirsky, Y.: Back-in-time diffusion: Unsu- pervised detection of medical deepfakes. ACM Transactions on Intelligent Systems and Technology16(6), 1–26 (2025)

    2025

  6. [6]

    In: European Conference on Com- puter Vision

    Hamamci, I.E., Er, S., Sekuboyina, A., Simsar, E., Tezcan, A., Simsek, A.G., Esirgun, S.N., Almas, F., Doğan, I., Dasdelen, M.F., et al.: Generatect: Text- conditional generation of 3d chest ct volumes. In: European Conference on Com- puter Vision. pp. 126–143. Springer (2024)

    2024

  7. [7]

    arXiv preprint arXiv:2403.17834 (2024)

    Hamamci, I.E., Er, S., Wang, C., Almas, F., Simsek, A.G., Esirgun, S.N., Do- gan, I., Durugol, O.F., Hou, B., Shit, S., et al.: Developing generalist foundation models from a multimodal dataset for 3d computed tomography. arXiv preprint arXiv:2403.17834 (2024)

    arXiv 2024

  8. [8]

    He,K.,Zhang,X.,Ren,S.,Sun,J.:Deepresiduallearningforimagerecognition.In: Proceedings of the IEEE conference on computer vision and pattern recognition. pp. 770–778 (2016)

    2016

Show all 30 references
  1. [9]

    arXiv preprint arXiv:2502.15204 (2025)

    Jiang, Y., Lemaréchal, Y., Bafaro, J., Abi-Rjeile, J., Joubert, P., Després, P., Manem, V.: Lung-ddpm: Semantic layout-guided diffusion models for thoracic ct image synthesis. arXiv preprint arXiv:2502.15204 (2025)

    2025 arXiv

  2. [10]

    Computers in biology and medicine 199, 111290 (2025)

    Jiang, Y., Shariftabrizi, A., Manem, V.S.: Lung-ddpm+: Efficient thoracic ct image synthesis using diffusion probabilistic model. Computers in biology and medicine 199, 111290 (2025)

    2025

  3. [11]

    In: International Conference on Medical Image Computing and Computer-Assisted Intervention

    Konz, N., Chen, Y., Dong, H., Mazurowski, M.A.: Anatomically-controllable med- ical image generation with segmentation-guided diffusion models. In: International Conference on Medical Image Computing and Computer-Assisted Intervention. pp. 88–98. Springer (2024)

    2024

  4. [12]

    Bioengineering9(11), 698 (2022)

    Lee, K.W., Chin, R.K.Y.: Diverse covid-19 ct image-to-image translation with stacked residual dropout. Bioengineering9(11), 698 (2022)

    2022

  5. [13]

    In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition

    Li, J., Xie, H., Li, J., Wang, Z., Zhang, Y.: Frequency-aware discriminative feature learning supervised by single-center loss for face forgery detection. In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition. pp. 6458–6467 (2021) 10 Y. Li et al

    2021

  6. [14]

    In: Pro- ceedings of the 31st ACM SIGKDD Conference on Knowledge Discovery and Data Mining V

    Li, O., Cai, J., Hao, Y., Jiang, X., Hu, Y., Feng, F.: Improving synthetic image detection towards generalization: An image transformation perspective. In: Pro- ceedings of the 31st ACM SIGKDD Conference on Knowledge Discovery and Data Mining V. 1. pp. 2405–2414 (2025)

    2025

  7. [15]

    Li,S.,Xing,Z.,Wang,H.,Hao,P.,Li,X.,Liu,Z.,Zhu,L.:Towardmedicaldeepfake detection:Acomprehensivedatasetandnovelmethod.In:InternationalConference on Medical Image Computing and Computer-Assisted Intervention. pp. 626–637. Springer (2025)

    2025

  8. [16]

    In: The Thirty-ninth Annual Conference on Neural Information Processing Systems (2025)

    Liang, S., Liu, J., Renzhang, C., Guan, Q.: Ferretnet: Efficient synthetic image detection via local pixel dependencies. In: The Thirty-ninth Annual Conference on Neural Information Processing Systems (2025)

    2025

  9. [17]

    Neural computing and applications37(35), 29099–29111 (2025)

    Loey, M., Manogaran, G., Khalifa, N.E.M.: A deep transfer learning model with classical data augmentation and cgan to detect covid-19 from chest ct radiography digital images. Neural computing and applications37(35), 29099–29111 (2025)

    2025

  10. [18]

    In: 28th USENIX Security Symposium (USENIX Security 19)

    Mirsky, Y., Mahler, T., Shelef, I., Elovici, Y.:{CT-GAN}: Malicious tampering of 3d medical imagery using deep learning. In: 28th USENIX Security Symposium (USENIX Security 19). pp. 461–478 (2019)

    2019

  11. [19]

    arXiv preprint arXiv:2506.00633 (2025)

    Molino, D., Caruso, C.M., Ruffini, F., Soda, P., Guarrasi, V.: Text-to-ct generation via 3d latent diffusion model with contrastive vision-language pretraining. arXiv preprint arXiv:2506.00633 (2025)

    2025

  12. [20]

    In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition

    Ojha, U., Li, Y., Lee, Y.J.: Towards universal fake image detectors that gener- alize across generative models. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. pp. 24480–24489 (2023)

    2023

  13. [21]

    Scientific Re- ports12(1), 18573 (2022)

    Prezja, F., Paloneva, J., Pölönen, I., Niinimäki, E., Äyrämö, S.: Deepfake knee osteoarthritis x-rays from generative adversarial neural networks deceive medical experts and offer augmentation potential to automatic classification. Scientific Re- ports12(1), 18573 (2022)

    2022

  14. [22]

    In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition

    Rombach, R., Blattmann, A., Lorenz, D., Esser, P., Ommer, B.: High-resolution image synthesis with latent diffusion models. In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition. pp. 10684–10695 (2022)

    2022

  15. [23]

    arXiv preprint arXiv:2011.13456 (2020)

    Song, Y., Sohl-Dickstein, J., Kingma, D.P., Kumar, A., Ermon, S., Poole, B.: Score- based generative modeling through stochastic differential equations. arXiv preprint arXiv:2011.13456 (2020)

    2011 arXiv

  16. [24]

    IEEE journal of biomedical and health informatics26(8), 3966–3975 (2022)

    Sun, L., Chen, J., Xu, Y., Gong, M., Yu, K., Batmanghelich, K.: Hierarchical amortized gan for 3d high resolution medical image synthesis. IEEE journal of biomedical and health informatics26(8), 3966–3975 (2022)

    2022

  17. [25]

    In: Proceedings of the AAAI Conference on Artificial Intelligence

    Tan, C., Zhao, Y., Wei, S., Gu, G., Liu, P., Wei, Y.: Frequency-aware deepfake detection: Improving generalizability through frequency space domain learning. In: Proceedings of the AAAI Conference on Artificial Intelligence. vol. 38, pp. 5052–5060 (2024)

    2024

  18. [26]

    In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition

    Tan, C., Zhao, Y., Wei, S., Gu, G., Liu, P., Wei, Y.: Rethinking 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. pp. 28130–28139 (2024)

    2024

  19. [27]

    In: Interna- tional conference on machine learning

    Tan, M., Le, Q.: Efficientnetv2: Smaller models and faster training. In: Interna- tional conference on machine learning. pp. 10096–10106. PMLR (2021)

    2021

  20. [28]

    IEEE Transactions on Medical Imag- ing43(10), 3648–3660 (2024) CTForensics 11

    Xu, Y., Sun, L., Peng, W., Jia, S., Morrison, K., Perer, A., Zandifar, A., Visweswaran, S., Eslami, M., Batmanghelich, K.: Medsyn: text-guided anatomy- aware synthesis of high-fidelity 3-d ct images. IEEE Transactions on Medical Imag- ing43(10), 3648–3660 (2024) CTForensics 11

    2024

  21. [29]

    In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition

    Yu, Z., Zhao, C., Wang, Z., Qin, Y., Su, Z., Li, X., Zhou, F., Zhao, G.: Searching central difference convolutional networks for face anti-spoofing. In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition. pp. 5295– 5305 (2020)

    2020

  22. [30]

    In: Proceedings of the IEEE/CVF International Conference on Computer Vision

    Zhang, T., Liu, P., Lu, Y., Cai, M., Zhang, Z., Zhang, Z., Zhou, Q.: Cwnet: Causal wavelet network for low-light image enhancement. In: Proceedings of the IEEE/CVF International Conference on Computer Vision. pp. 8789–8799 (2025)

    2025

Pith tools

Reviewed August 2, 2026 · model on record in the stance chip above.