Pith. sign in

REVIEW 4 major objections 6 minor 2 cited by

ArtBrain: An Explainable end-to-end Toolkit for Classification and Attribution of AI-Generated Art and Style

T0 review · 4 major / 6 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 The paper claims that a convolutional network, trained on a new 185,015-image dataset of human and diffusion-generated art, can classify art style and source with F1 0.869 and attribute the generating model with 0.999 accuracy…

desk verdict Useful dataset and toolkit, but the headline accuracy numbers are not trustworthy because the test set was used as the validation set for checkpoint selection. read the letter →

arxiv 2412.01512 v1 pith:45BTLSRC submitted 2024-12-02 cs.AI cs.CV

Ravidu Suien Rammuni Silva , Ahmad Lotfi , Isibor Kennedy Ihianle , Golnaz Shahtahmassebi , Jordan J. Bird This is my paper · ORCID
classification cs.AIcs.CV
keywords AIartdetectionattributionAI-ArtBenchdatasetdiffusionmodelsAttentionConvNeXtexplainableArtisticTuringTestconvolutionalneuralnetworks
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

This paper aims to give researchers and the public a working way to tell AI-generated artwork apart from human-made art and to name the model that produced it. To that end, it introduces AI-ArtBench, a dataset of 185,015 images in ten art styles: 125,015 images generated by two diffusion models (Latent Diffusion and Stable Diffusion) plus 60,000 human artworks. It then proposes AttentionConvNeXt, a convolutional network that reports an F1-score of 0.869 for jointly classifying style and source, and an attribution accuracy of 0.999 for identifying the generating model. In an Artistic Turing Test with 50 participants, humans identified AI images with roughly 54–58% accuracy while the model scored 98% on the same images, leading the authors to claim that diffusion-generated artworks are better identified and described by CNNs than by humans.

What carries the argument

The load-bearing object is AttentionConvNeXt: a ConvNeXt CNN whose low- and mid-level feature blocks are frozen after ImageNet pretraining, while its high-level blocks are fine-tuned and an attention module (built from squeeze-and-excitation blocks) concatenates the feature maps from all three depths, computes per-channel importance via two fully connected layers, and reweights the concatenated maps before a two-layer classifier. This design is what lets the model use low-, mid-, and high-level visual features together; the ablation shows it adds about 3% F1 over plain ConvNeXt with transfer learning, and the attribution head (the same classifier's source labels) is what yields the near-perfect 0.999 source attribution.

What would settle it

Run the trained AttentionConvNeXt on a fresh 10,000-image set of art from the same ten styles and same three sources that was never used for checkpoint selection or early stopping; if the 30-class F1 drops substantially below 0.869 or the attribution accuracy falls below 0.999, the reported generalization is not as strong as claimed. A second check is to feed images from a diffusion model outside the training set, such as SDXL or Midjourney, and see whether attribution accuracy stays near ceiling or collapses.

Watch

Extended reading notes

Core claim

The central claim is that a single CNN, AttentionConvNeXt, can both tell whether an artwork is human-made or machine-generated and attribute it to the specific diffusion model that produced it. On the AI-ArtBench test split, the authors report an F1-score of 0.869 for the 30-way classification of style plus source, and an attribution accuracy of 0.999 across the three sources (human, Latent Diffusion, Stable Diffusion). Presented in good faith, the paper's conclusion goes further: because the model reached 98% on the same 50 images on which human participants averaged around 54% in the Artistic Turing Test, the authors state that AI diffusion-generated artworks are better identified and described by CNNs than by humans. The attribution result is interpreted as evidence that diffusion models leave a detectable statistical fingerprint in their outputs, extending earlier attribution work from general images to fine art.

Load-bearing premise

The load-bearing premise is that the 10,000-image test split gives an unbiased estimate of how the model will do on new art, even though the same split was used to pick the best training epoch and to report accuracy.

Editorial extensions

If this is right

  • If the numbers hold, ArtBrain gives competition organizers, galleries, and platforms a practical tool to screen submissions for AI generation and to name the likely model (Latent vs Stable Diffusion).
  • The near-perfect attribution accuracy suggests image generators leave identifiable statistical fingerprints, so attribution is feasible even when metadata is stripped.
  • The dataset's size (185,015 images across 10 styles) provides a benchmark for future detectors and generators, filling the gap the paper identifies in available AI-art datasets.
  • The Turing Test result implies that automated detection is substantially more reliable than human judgment on diffusion-art images, supporting the paper's conclusion.
  • The FM-G-CAM heatmaps add multi-class explanations, so users can see which visual structures drove a style/source decision rather than just a label.

Reading between the lines

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

  • Because the same images were used for validation and test, the headline F1 and attribution numbers are likely optimistic; a properly separated hold-out would be expected to show lower performance.
  • The contrast and photo-replication experiments suggest the detector is sensitive to low-level image transforms; this implies that in real use, simple edits could fool attribution, so robustness to perturbations is a natural next test.
  • If attribution fingerprints are model-specific, the method may transfer to newer diffusion models (e.g., SDXL or Midjourney) only if those models share traceable artifacts; the paper's own tests on one Midjourney piece hint at this but do not prove it.
  • The human-model comparison should be read with the small participant pool in mind (50 respondents, mostly self-reported non-experts); a larger, more diverse sample could change the human baseline.
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

4 major / 6 minor

Summary. The paper introduces AI-ArtBench, a large dataset of 185,015 artistic images (125,015 AI-generated by Latent Diffusion and Stable Diffusion, 60,000 human artworks), and proposes AttentionConvNeXt, a CNN architecture combining ConvNeXt with a multi-level attention module, for classifying art style and attributing the generative source. The authors also present ArtBrain, a web-based toolkit, and an Artistic Turing Test with 50 participants. The paper claims an F1-score of 0.869 for the 30-class style/source classification, an attribution accuracy of 0.999, and that the CNN outperforms humans (98% vs. about 54%) at distinguishing AI-generated from human art. The dataset and code are publicly released.

Significance. If the reported numbers were reliable, this would be a useful contribution: it is one of the few large-scale datasets for AI-art detection and attribution, the toolkit is practical, and the architectural idea of combining low/mid/high-level features with squeeze-and-excitation across feature blocks is interesting. The open release of the dataset and model code is a clear strength. However, the central empirical claims are undermined by the evaluation protocol, which uses the test set for model selection, and the human-comparison conclusion rests on that biased evaluation. The paper's own robustness tests also raise questions about the practical utility of the model for real-world digitized or photographed artworks.

major comments (4)
  1. [§4.2 and §5] The evaluation is invalid because the test set is used for model selection. Section 4.2 states 'The testing set was used as the Validation set in the training' and the best epoch (epoch 15) was chosen by lowest validation loss on that set; Section 5 then states 'The same set used for validation is used as the test set.' Selecting the checkpoint on the same images that later produce the reported F1 (0.869) and attribution accuracy (0.999) biases those numbers upward in an unknown amount, and the comparison with human performance in Section 5.5 inherits this bias. The authors must retrain with a proper train/validation/test split and report metrics on a genuinely held-out test set.
  2. [Abstract vs. Table 5] The abstract reports human accuracy of approximately 58% and AI accuracy of around 99%, but Table 5 reports overall human accuracy of 53.8% and ArtBrain accuracy of 98% on the same set. These numbers should be reconciled; the discrepancy could indicate an error in the abstract or in the table, and in either case the reader cannot determine which value is authoritative.
  3. [§5.4] The contrast and re-photography tests show severe prediction instability: changing contrast by ±100% flips the top prediction for 'Starry Night' from Post-Impressionism-Human to other classes, and capturing a photo of the same image on a 2K screen changes the top prediction from 84% Post-Impressionism-Human to 73% Expressionism-Human. These results are acknowledged as limitations, but they undercut the claim that ArtBrain is a 'ready-to-use' detector for real-world art, where scanning and photographing are common. The conclusion that CNNs are better than humans at identifying AI art should be tempered by this operational fragility.
  4. [§5.5] The Artistic Turing Test is underpowered and its comparison is biased. Only 50 participants and 50 images were used, and the AI's 98% accuracy is computed on the same test-as-validation set that was used to select the model, while the human accuracy is on the same images but without any model-selection bias. The claim that 'AI diffusion-generated artworks are better identified and described by CNNs than by humans' is not supported by the reported protocol; a fair comparison requires an unbiased test set and a larger, more balanced participant sample.
minor comments (6)
  1. [§4.1] Please fix the typo 'Intel Xenon' to 'Intel Xeon' in the hardware description.
  2. [§5.5] The phrase 'resting the practical effectiveness' should read 'testing the practical effectiveness'.
  3. [Table 3] The table header says 'F1-Score' but the reported values are attribution accuracies; please rename the column to 'Accuracy' or clarify whether F1 is actually meant.
  4. [§2.3] The reference to 'CIFAKE[?]' is unresolved and should be completed or removed.
  5. [§5.1 and Table 2] Consider adding a column description for 'ArtBench' as a baseline; the current table does not indicate whether the ArtBench column refers to a pretrained model or the original dataset's reported score.
  6. [Figure 5] The visual difference between the preprocessed and original image is not clear from the figure; adding a legend or a quantitative description of preprocessing (e.g., normalization values) would help.

Circularity Check

0 steps flagged · score 2.0 of 10

No circular derivation in the central claims; the accuracy numbers are empirical results, with a validation/test-set overlap that is a validity concern rather than a circularity.

full rationale

The paper's main claims are empirical measurements: a CNN is trained on a dataset and then evaluated on a held-out split. No equation-level reduction makes a predicted quantity equal to an input quantity by definition, and no fitted parameter is renamed as a prediction. The AttentionConvNeXt architecture is a concrete combination of ConvNeXt and SE-style attention, trained with standard supervised learning; the reported F1 and attribution scores are model outputs, not identities. The self-citations to FM-G-CAM [34] and ParallelXNet [48] are disclosed design references for the attention module and for explainability, and they are not used to define the accuracy numbers or to rule out alternative methods, so they are not load-bearing circularity. The paper also compares against MobileNetV2 and ArtBench baselines on the same test set, giving it internal empirical context. One flagged concern is that Section 4.2 states 'The testing set was used as the Validation set in the training,' and Section 5 states 'The same set used for validation is used as the test set,' meaning the checkpoint (epoch 15) was chosen using the same images later used for the reported metrics. This can inflate the reported accuracy and weakens the human-comparison conclusion, but it is a methodological leakage/validity issue rather than a circular derivation: the model's predictions are not constructed from the test labels, and the reported metric is not identical to a fitted parameter by construction. The paper's own limitations acknowledge dataset variety, Turing-test size, and exploitation-protection gaps, none of which constitute circularity. Overall, the derivation chain is self-contained and not self-referential, so the score is low, reflecting only minor non-load-bearing self-citation and the flagged validation-overlap concern.

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

The central accuracy claims rest on several hand-chosen hyperparameters, on dataset-generation choices, and on the assumption that the test split can double as a validation split without bias. That last assumption is explicit in the paper and is violated by the protocol.

free parameters (5)
  • Learning rate and schedule = 0.001 initial, reduced by factor 10 after 2 epochs without improvement
    Selected via optimization experiments on the validation set, which is the same split later reported as the test set.
  • Number of training epochs and checkpoint epoch = 18 epochs; epoch 15 selected
    Early stopping was based on validation loss on the test split, biasing the reported accuracy.
  • Batch size = 32
    Reported as a chosen hyperparameter; no systematic search described.
  • SE attention reduction factor mu = not reported
    The 'squeeze and excitation' feature reduction factor in Equation 6 is a model capacity choice that is not specified.
  • Diffusion generation settings = 50 steps, guidance scale 9 (SD), diversity scale 5.0 (LD), negative prompt 'photo frame' (SD)
    Hand-chosen generation parameters shape the distribution of AI images in the dataset and thus the attribution fingerprints the model learns.
assumptions (4)
  • domain assumption The 60,000 ArtBench human images adequately represent human art across the ten styles.
    The human class of AI-ArtBench is inherited from ArtBench (Section 4.1) without additional filtering or augmentation.
  • domain assumption Two diffusion models, Latent Diffusion and Stable Diffusion, sufficiently represent the space of AI-generated art.
    All 125,015 AI images were generated with these two models; Section 6 lists limited generative models as a limitation.
  • domain assumption A learnable source fingerprint exists in AI-generated images, as suggested by Sha et al.
    Section 6 invokes the fingerprint idea to explain the 0.999 attribution accuracy.
  • ad hoc to paper The test split can be used for validation and model selection without biasing the reported evaluation.
    Section 4.2 states 'The testing set was used as the Validation set in the training' and Section 5 states 'The same set used for validation is used as the test set'; this is the protocol the accuracy claims depend on.

how reviews work

0 comments
Cite this review

Pith. "Pith review of ArtBrain: An Explainable end-to-end Toolkit for Classification and Attribution of AI-Generated Art and Style." pith.science (2026). https://pith.science/paper/45BTLSRC

@misc{pith2026241201512,
  author       = {Pith},
  title        = {Pith review of: ArtBrain: An Explainable end-to-end Toolkit for Classification and Attribution of AI-Generated Art and Style},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/45BTLSRC}},
  note         = {Machine review of arXiv:2412.01512}
}
read the original abstract

Recently, the quality of artworks generated using Artificial Intelligence (AI) has increased significantly, resulting in growing difficulties in detecting synthetic artworks. However, limited studies have been conducted on identifying the authenticity of synthetic artworks and their source. This paper introduces AI-ArtBench, a dataset featuring 185,015 artistic images across 10 art styles. It includes 125,015 AI-generated images and 60,000 pieces of human-created artwork. This paper also outlines a method to accurately detect AI-generated images and trace them to their source model. This work proposes a novel Convolutional Neural Network model based on the ConvNeXt model called AttentionConvNeXt. AttentionConvNeXt was implemented and trained to differentiate between the source of the artwork and its style with an F1-Score of 0.869. The accuracy of attribution to the generative model reaches 0.999. To combine the scientific contributions arising from this study, a web-based application named ArtBrain was developed to enable both technical and non-technical users to interact with the model. Finally, this study presents the results of an Artistic Turing Test conducted with 50 participants. The findings reveal that humans could identify AI-generated images with an accuracy of approximately 58%, while the model itself achieved a significantly higher accuracy of around 99%.

Figures

Figures reproduced from arXiv: 2412.01512 by the authors.

Figure 1
Figure 1. ‘AttentionConvNeXt’ model architecture design. [PITH_FULL_IMAGE:figures/full_fig_p005_1.png] view at source ↗
Figure 2
Figure 2. Attention Module Architecture. was initially loaded with the weights trained on ImageNet and further fine-tuned in training. ‘Attention Module’ receives inputs from all the feature blocks mentioned above where all the blocks are concatenated into one deep feature map block, which is then weighted using the concept of ‘squeeze and excitation’ [46]. ‘Classifier FCN’ receives the weighted feature maps from the Attentio… view at source ↗
Figure 3
Figure 3. Image counts in each class [PITH_FULL_IMAGE:figures/full_fig_p007_3.png] view at source ↗
Figures from the paper (8 more)
Figure 4
Figure 4. Figure 4: Sample of the dataset representing each style generated using each source. Rows (top to bottom): Latent [PITH_FULL_IMAGE:figures/full_fig_p007_4.png]
Figure 5
Figure 5. Figure 5: Preprocessed image Vs Original image (Starry Night by Vincent-van-Gogh, 1889). [PITH_FULL_IMAGE:figures/full_fig_p008_5.png]
Figure 6
Figure 6. Figure 6: Prediction results on “Starry Nights" with different contrast levels. Left: -100% Contrast, Middle: Reference, [PITH_FULL_IMAGE:figures/full_fig_p012_6.png]
Figure 7
Figure 7. Figure 7: Prediction results of “Starry Nights" (top row) and its replication (bottom row). [PITH_FULL_IMAGE:figures/full_fig_p012_7.png]
Figure 8
Figure 8. Figure 8: Initial Question of ‘Artistic Turing Test’. [PITH_FULL_IMAGE:figures/full_fig_p013_8.png]
Figure 9
Figure 9. Figure 9: ‘Artistic Turing Test’ Question example. This is an AI-generated art which 84% of the Artistic Turing Test [PITH_FULL_IMAGE:figures/full_fig_p013_9.png]
Figure 10
Figure 10. Figure 10: The only incorrect prediction made by the ArtBrain model and its accompanying saliency map. The model [PITH_FULL_IMAGE:figures/full_fig_p014_10.png]
Figure 11
Figure 11. Figure 11: Grad-CAM Vs. FM-G-CAM Comparison. AI-ArtBench, a sufficiently sized dataset with more than 120,000 AI-generated artworks and 60,000 real art images from the ArtBench dataset, totalling 180,000+ images altogether (see [PITH_FULL_IMAGE:figures/full_fig_p015_11.png]

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 2 Pith papers

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

  1. Identifying Prompted Artist Names from Generated Images

    cs.CV 2025-07 conditional novelty 7.0 of 10

    A new 1.95M-image benchmark measures how well vision models identify artist names explicitly prompted into text-to-image systems, across artists, prompts, generators, and artist counts.

  2. Robustness of AI-Art Detectors under Generator Shift

    cs.CV 2026-08 conditional novelty 5.0 of 10

    AI-art detectors trained on LDM and SD2.1 artwork miss 42% to 58% of Stable Diffusion 3.5 Medium images in a zero-shot cross-generator test.

Reference graph

Works this paper leans on

53 extracted references · 32 canonical work pages · cited by 2 Pith papers

  1. [1]

    Podell, Z

    D. Podell, Z. English, K. Lacey, A. Blattmann, T. Dockhorn, J. Müller, J. Penna, R. Rombach, SDXL: Improving Latent Diffusion Models for High-Resolution Image Synthesis, publisher: arXiv Version Number: 1 (2023). 16AI-ArtBench dataset available here: https://www.kaggle.com/datasets/ravidussilva/real-ai-art 16 Explainable Classification and Attribution of ...

    arXiv doi:10.48550/arxiv.2307.01952 2023

  2. [2]

    Rombach, A

    R. Rombach, A. Blattmann, D. Lorenz, P. Esser, B. Ommer, High-Resolution Image Synthesis with Latent Diffusion Models, in: 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), IEEE, New Orleans, LA, USA, 2022, pp. 10674–10685. doi:10.1109/CVPR52688.2022.01042

    arXiv 2022

  3. [3]

    Ramesh, P

    A. Ramesh, P. Dhariwal, A. Nichol, C. Chu, M. Chen, Hierarchical Text-Conditional Image Generation with CLIP Latents, publisher: arXiv Version Number: 1 (2022). doi:10.48550/ARXIV.2204.06125

    arXiv doi:10.48550/arxiv.2204.06125 2022

  4. [4]

    Roose, An a.i.-generated picture won an art prize

    K. Roose, An a.i.-generated picture won an art prize. artists aren’t happy., The New York Times (09 2022). URL https://www.nytimes.com/2022/09/02/technology/ai-artificial-intelligence-artists .html

    work page 2022

  5. [5]

    Grierson, Photographer admits prize-winning image was ai-generated, The Guardian (04 2023)

    J. Grierson, Photographer admits prize-winning image was ai-generated, The Guardian (04 2023). URL https://www.theguardian.com/technology/2023/apr/17/photographer-admits-prize-win ning-image-was-ai-generated

    work page 2023

  6. [6]

    Feller, On the Theory of Stochastic Processes, with Particular Reference to Applications, in: First Berkeley Symposium on Mathematical Statistics and Probability, 1949, pp

    W. Feller, On the Theory of Stochastic Processes, with Particular Reference to Applications, in: First Berkeley Symposium on Mathematical Statistics and Probability, 1949, pp. 403–432

    work page 1949

  7. [7]

    Goodfellow, J

    I. Goodfellow, J. Pouget-Abadie, M. Mirza, B. Xu, D. Warde-Farley, S. Ozair, A. Courville, Y . Bengio, Generative Adversarial Nets, in: Z. Ghahramani, M. Welling, C. Cortes, N. Lawrence, K. Q. Weinberger (Eds.), Advances in Neural Information Processing Systems, V ol. 27, Curran Associates, Inc., 2014

    work page 2014

  8. [8]

    D. P. Kingma, M. Welling, Auto-Encoding Variational Bayes, in: 2nd International Conference on Learning Representations, ICLR 2014, Banff, AB, Canada, April 14-16, 2014, Conference Track Proceedings, 2014

    work page 2014

Show all 53 references
  1. [9]

    Karras, S

    T. Karras, S. Laine, T. Aila, A Style-Based Generator Architecture for Generative Adversarial Networks, in: 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), IEEE, Long Beach, CA, USA, 2019, pp. 4396–4405. doi:10.1109/CVPR.2019.00453

    2019

  2. [10]

    Sohl-Dickstein, E

    J. Sohl-Dickstein, E. Weiss, N. Maheswaranathan, S. Ganguli, Deep Unsupervised Learning using Nonequilibrium Thermodynamics, in: F. Bach, D. Blei (Eds.), Proceedings of the 32nd International Conference on Machine Learning, V ol. 37 of Proceedings of Machine Learning Research,...

    2015

  3. [11]

    J. Ho, A. Jain, P. Abbeel, Denoising Diffusion Probabilistic Models, in: H. Larochelle, M. Ranzato, R. Hadsell, M. F. Balcan, H. Lin (Eds.), Advances in Neural Information Processing Systems, V ol. 33, Curran Associates, Inc., 2020, pp. 6840–6851

    2020

  4. [12]

    Dhariwal, A

    P. Dhariwal, A. Nichol, Diffusion Models Beat GANs on Image Synthesis, in: M. Ranzato, A. Beygelzimer, Y . Dauphin, P. S. Liang, J. W. Vaughan (Eds.), Advances in Neural Information Processing Systems, V ol. 34, Curran Associates, Inc., 2021, pp. 8780–8794

    2021

  5. [13]

    Saharia, W

    C. Saharia, W. Chan, S. Saxena, L. Li, J. Whang, E. L. Denton, K. Ghasemipour, R. Gontijo Lopes, B. Karagol Ayan, T. Salimans, J. Ho, D. J. Fleet, M. Norouzi, Photorealistic Text-to-Image Diffusion Models with Deep Language Understanding, in: S. Koyejo, S. Mohamed, A. Agarwal,...

    2022

  6. [14]

    H. Qiao, V . Liu, L. Chilton, Initial Images: Using Image Prompts to Improve Subject Representation in Multimodal AI Generated Art, in: Creativity and Cognition, ACM, Venice Italy, 2022, pp. 15–28. doi:10.1145/3527927. 3532792

    2022 doi

  7. [15]

    Epstein, S

    Z. Epstein, S. Levine, D. G. Rand, I. Rahwan, Who Gets Credit for AI-Generated Art?, iScience 23 (9) (2020) 101515. doi:10.1016/j.isci.2020.101515

    2020

  8. [16]

    Z. Liu, H. Mao, C.-Y . Wu, C. Feichtenhofer, T. Darrell, S. Xie, A convnet for the 2020s, in: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2022, pp. 11976–11986

    2022

  9. [17]

    Vaswani, N

    A. Vaswani, N. Shazeer, N. Parmar, J. Uszkoreit, L. Jones, A. N. Gomez, L. Kaiser, I. Polosukhin, Attention is All you Need, in: I. Guyon, U. V . Luxburg, S. Bengio, H. Wallach, R. Fergus, S. Vishwanathan, R. Garnett (Eds.), Advances in Neural Information Processing Systems, V...

    2017

  10. [18]

    Ganteführer-Trier, Cubism, Taschen, 2004

    A. Ganteführer-Trier, Cubism, Taschen, 2004

    2004

  11. [19]

    Nadeau, The History of Surrealism, New York: Macmillan, 1965

    M. Nadeau, The History of Surrealism, New York: Macmillan, 1965

    1965

  12. [20]

    R. L. Herbert, Impressionism: art, leisure, and Parisian society, Yale University Press, 1988

    1988

  13. [21]

    Honour, Romanticism, Routledge, 2018

    H. Honour, Romanticism, Routledge, 2018

    2018

  14. [22]

    Hornstein, J

    N. Hornstein, J. Nunes, K. K. Grohmann, Understanding minimalism, Cambridge University Press, 2005

    2005

  15. [23]

    N. Wolf, U. Grosenick, Expressionism, Taschen, 2004. 17 Explainable Classification and Attribution of AI-Generated Art and Style Silva et al

    2004

  16. [24]

    Harris, Ukiyo-e: the art of the Japanese print, Tuttle Publishing, 2012

    F. Harris, Ukiyo-e: the art of the Japanese print, Tuttle Publishing, 2012

    2012

  17. [25]

    Strezoski, M

    G. Strezoski, M. Worring, OmniArt: A Large-scale Artistic Benchmark, ACM Transactions on Multimedia Computing, Communications, and Applications 14 (4) (2018) 1–21. doi:10.1145/3273022

    2018 doi

  18. [26]

    H. Mao, M. Cheung, J. She, DeepArt: Learning Joint Representations of Visual Arts, in: Proceedings of the 25th ACM international conference on Multimedia, ACM, Mountain View California USA, 2017, pp. 1183–1191. doi:10.1145/3123266.3123405

    2017

  19. [27]

    Mensink, J

    T. Mensink, J. Van Gemert, The Rijksmuseum Challenge: Museum-Centered Visual Recognition, in: Proceedings of International Conference on Multimedia Retrieval, ACM, Glasgow United Kingdom, 2014, pp. 451–454. doi:10.1145/2578726.2578791

    2014

  20. [28]

    P. Liao, X. Li, X. Liu, K. Keutzer, The ArtBench Dataset: Benchmarking Generative Models with Artworks (2022)

    2022

  21. [29]

    Garcia, C

    N. Garcia, C. Ye, Z. Liu, Q. Hu, M. Otani, C. Chu, Y . Nakashima, T. Mitamura, A Dataset and Baselines for Visual Question Answering on Art, in: Proceedings of the European Conference in Computer Vision Workshops, 2020

    2020

  22. [30]

    Mohammad, S

    S. Mohammad, S. Kiritchenko, Wikiart emotions: An annotated dataset of emotions evoked by art, in: Proceedings of the eleventh international conference on language resources and evaluation (LREC 2018), 2018

    2018

  23. [31]

    S. A. Amirshahi, G. U. Hayn-Leichsenring, J. Denzler, C. Redies, JenAesthetics Subjective Dataset: Analyzing Paintings by Subjective Scores, in: L. Agapito, M. M. Bronstein, C. Rother (Eds.), Computer Vision - ECCV 2014 Workshops, V ol. 8925, Springer International Publishing,...

    2014 doi

  24. [32]

    LeCun, Y

    Y . LeCun, Y . Bengio, G. Hinton, Deep learning, Nature 521 (7553) (2015) 436–444.doi:10.1038/nature14 539

    2015 doi

  25. [33]

    R. R. Selvaraju, M. Cogswell, A. Das, R. Vedantam, D. Parikh, D. Batra, Grad-CAM: Visual Explanations from Deep Networks via Gradient-Based Localization, in: 2017 IEEE International Conference on Computer Vision (ICCV), IEEE, Venice, 2017, pp. 618–626. doi:10.1109/ICCV.2017.74

    2017 doi

  26. [34]

    R. S. R. Silva, J. J. Bird, Fm-g-cam: A holistic approach for explainable ai in computer vision (2023). arXiv: 2312.05975

    2023 arXiv

  27. [35]

    A. Hertzmann, Painterly rendering with curved brush strokes of multiple sizes, in: Proceedings of the 25th annual conference on Computer graphics and interactive techniques - SIGGRAPH ’98, ACM Press, Not Known, 1998, pp. 453–460. doi:10.1145/280814.280951

    1998

  28. [36]

    J.-Y . Zhu, T. Park, P. Isola, A. A. Efros, Unpaired Image-to-Image Translation Using Cycle-Consistent Adversarial Networks, in: 2017 IEEE International Conference on Computer Vision (ICCV), IEEE, Venice, 2017, pp. 2242–2251. doi:10.1109/ICCV.2017.244

    2017 doi

  29. [37]

    L. A. Gatys, A. S. Ecker, M. Bethge, Image Style Transfer Using Convolutional Neural Networks, in: 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), IEEE, Las Vegas, NV , USA, 2016, pp. 2414–2423. doi:10.1109/CVPR.2016.265

    2016 doi

  30. [38]

    Y . Song, S. Ermon, Generative modeling by estimating gradients of the data distribution, Advances in neural information processing systems 32 (2019)

    2019

  31. [39]

    Hyvärinen, P

    A. Hyvärinen, P. Dayan, Estimation of non-normalized statistical models by score matching., Journal of Machine Learning Research 6 (4) (2005)

    2005

  32. [40]

    X. Wu, B. R. Brooks, Self-guided Langevin dynamics simulation method, Chemical Physics Letters 381 (3-4) (2003) 512–518. doi:10.1016/j.cplett.2003.10.013

    2003 doi

  33. [41]

    S. Gao, P. Zhou, M.-M. Cheng, S. Yan, Masked Diffusion Transformer is a Strong Image Synthesizer, publisher: arXiv Version Number: 1 (2023). doi:10.48550/ARXIV.2303.14389

    2023 doi arXiv

  34. [42]

    Y . Song, P. Dhariwal, M. Chen, I. Sutskever, Consistency Models, in: A. Krause, E. Brunskill, K. Cho, B. Engel- hardt, S. Sabato, J. Scarlett (Eds.), Proceedings of the 40th International Conference on Machine Learning, V ol. 202 of Proceedings of Machine Learning Research, P...

    2023

  35. [43]

    Rossler, D

    A. Rossler, D. Cozzolino, L. Verdoliva, C. Riess, J. Thies, M. Niessner, FaceForensics++: Learning to Detect Manipulated Facial Images, in: Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV), 2019

    2019

  36. [44]

    Wolter, F

    M. Wolter, F. Blanke, R. Heese, J. Garcke, Wavelet-packets for deepfake image analysis and detection, Machine Learning 111 (11) (2022) 4295–4327. doi:10.1007/s10994-022-06225-5 . 18 Explainable Classification and Attribution of AI-Generated Art and Style Silva et al

    2022 doi

  37. [45]

    Dogoulis, G

    P. Dogoulis, G. Kordopatis-Zilos, I. Kompatsiaris, S. Papadopoulos, Improving Synthetically Generated Image Detection in Cross-Concept Settings, in: Proceedings of the 2nd ACM International Workshop on Multimedia AI against Disinformation, ACM, Thessaloniki Greece, 2023, pp. 2...

    2023

  38. [46]

    J. Hu, L. Shen, G. Sun, Squeeze-and-Excitation Networks, in: 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition, IEEE, Salt Lake City, UT, 2018, pp. 7132–7141. doi:10.1109/CVPR.2018.00745

    2018

  39. [47]

    J. Deng, W. Dong, R. Socher, L.-J. Li, Kai Li, Li Fei-Fei, ImageNet: A large-scale hierarchical image database, in: 2009 IEEE Conference on Computer Vision and Pattern Recognition, IEEE, Miami, FL, 2009, pp. 248–255. doi:10.1109/CVPR.2009.5206848

    2009

  40. [48]

    R. S. Rammuni Silva, P. Fernando, Effective Utilization of Multiple Convolutional Neural Networks for Chest X-Ray Classification, SN Computer Science 3 (6) (2022) 492. doi:10.1007/s42979-022-01390-9

    2022 doi

  41. [49]

    Sandler, A

    M. Sandler, A. Howard, M. Zhu, A. Zhmoginov, L.-C. Chen, MobileNetV2: Inverted Residuals and Linear Bottlenecks, in: 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition, IEEE, Salt Lake City, UT, 2018, pp. 4510–4520. doi:10.1109/CVPR.2018.00474

    2018

  42. [50]

    A. M. Turing, Computing machinery and intelligence, Springer, 2009

    2009

  43. [51]

    What Is Human?

    A. Daniele, C. Di Bernardi Luft, N. Bryan-Kinns, “What Is Human?” A Turing Test for Artistic Creativity, in: J. Romero, T. Martins, N. Rodríguez-Fernández (Eds.), Artificial Intelligence in Music, Sound, Art and Design, V ol. 12693, Springer International Publishing, Cham, 202...

    2021 doi

  44. [52]

    Z. Sha, Z. Li, N. Yu, Y . Zhang, DE-FAKE: Detection and Attribution of Fake Images Generated by Text-to-Image Generation Models, publisher: arXiv Version Number: 2 (2022). doi:10.48550/ARXIV.2210.06998

    2022 doi arXiv

  45. [53]

    Mona Lisa

    E. Collins, R. Achanta, S. Susstrunk, Deep feature factorization for concept discovery, in: Proceedings of the European Conference on Computer Vision (ECCV), 2018, pp. 336–352. 19 Explainable Classification and Attribution of AI-Generated Art and Style Silva et al. A ArtBrain ...

    2018

Pith tools

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