The DOI matches its Crossref record; 21 of the 21 references it cites were found in a registry or an index.
Detecting and Simulating Artifacts in GAN Fake Images
Open accessAbstract
To detect GAN generated images, conventional supervised machine learning algorithms require collecting a large number of real images as well as fake images generated by the targeted GAN model. However, the specific model used by the attacker is often unavailable. To address this, we propose a GAN simulator, AutoGAN, which can simulate the artifacts produced by the common pipeline shared by several popular GAN models. Additionally, we identify a unique artifact caused by the up-sampling component included in the common GAN pipelines. We show theoretically such artifacts are manifested as replications of spectra in the frequency domain and thus propose a classifier model based on the spectrum input, rather than the pixel input. By using the simulated images to train a spectrum based classifier, even without seeing the fake images produced by the targeted GAN model during training, our approach achieves state-of-the-art performances on detecting fake images generated by popular GAN models such as CycleGAN.
Overview
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
In the citing paper
“Figure 7 represents the potentials of employing the frequency artifacts for detecting DeepFakes, where the GAN-based manipulation introduces invisible artifacts in the frequency domain.”
From Countering Malicious DeepFakes: Survey, Battleground, and Horizon (2022) · cites this paper as Zhang et al. 2019
In the citing paper
“Due to limited data, network structure,8 and optimization algorithms,6 the GANs can hardly fit the distribution of real data.”
From Deepfakes: A new threat to image fabrication in scientific publications? (2022) · cites this paper as [8]
In the citing paper
“The most well-known artifact is the checkerboard effect due to the image upsampling operation in GAN.8”
From Deepfakes: A new threat to image fabrication in scientific publications? (2022) · cites this paper as [8]
Citations
Cites
Unpaired image-to-image translation using cycle-consistent adversarial networks
Zhu, J.-Y., Park, T., Isola, P., & Efros, A. A. (2017) · 2017 IEEE International Conference on Computer Vision (ICCV) · doi:10.1109/iccv.2017.244 · Cited by 23,602
Image-to-Image Translation with Conditional Adversarial Networks
Isola, P., Zhu, J.-Y., Zhou, T., & Efros, A. A. (2017) · 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2017.632 · Cited by 22,930 · cited 1×
“image that is constrained by the layout of the sketch [11], [6], [10].”
BackgroundLabaca-Castro, R. (2023) · Springer eBooks · doi:10.1007/978-3-658-40442-0_9 · Cited by 21,119
GAN（Generative Adversarial Nets）
淳司, 柴. (2017) · Journal of Japan Society for Fuzzy Theory and Intelligent Informatics · doi:10.3156/jsoft.29.5_177_2 · Cited by 20,507
Unsupervised representation learning with deep convolutional generative adversarial networks
Radford, A., Metz, L., & Chintala, S. (2015) · arXiv · doi:10.48550/arxiv.1511.06434 · Cited by 12,235 · cited 1×
“a deconvolution) [19], [6], [7] and nearest neighbor interpolation [9], [10].”
BackgroundA Style-Based Generator Architecture for Generative Adversarial Networks
Karras, T., Laine, S., & Aila, T. (2019) · 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2019.00453 · Cited by 11,714 · cited 2×
“In image generation, GAN [1] can be applied in the following scenarios: 1) taking noise as input to produce an image [9], [2], [3]; 2) taking an image from one semantic category (such as horse) as input to produce an image of another semantic category (such as zebra) [6], [7], [10]; 3) taking a sketch or a pixel level…”
BackgroundMethodProgressive Growing of GANs for Improved Quality, Stability, and Variation
Karras, T., Aila, T., Laine, S., & Lehtinen, J. (2017) · arXiv · doi:10.48550/arxiv.1710.10196 · Cited by 4,727 · cited 4×
“) [6] and nearest neighbor interpolation (NN) [9].”
BackgroundMethodInfluentialStarGAN: Unified Generative Adversarial Networks for Multi-domain Image-to-Image Translation
Choi, Y., Choi, M., Kim, M., Ha, J.-W., Kim, S., & Choo, J. (2018) · 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition · doi:10.1109/cvpr.2018.00916 · Cited by 4,163
Large Scale GAN Training for High Fidelity Natural Image Synthesis
Brock, A., Donahue, J., & Simonyan, K. (2018) · arXiv · doi:10.48550/arxiv.1809.11096 · Cited by 3,717 · cited 2×
“In image generation, GAN [1] can be applied in the following scenarios: 1) taking noise as input to produce an image [9], [2], [3]; 2) taking an image from one semantic category (such as horse) as input to produce an image of another semantic category (such as zebra) [6], [7], [10]; 3) taking a sketch or a pixel level…”
BackgroundMethodSemantic Image Synthesis With Spatially-Adaptive Normalization
Park, T., Liu, M.-Y., Wang, T.-C., & Zhu, J.-Y. (2019) · 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2019.00244 · Cited by 2,974 · cited 5×
“However, all classifiers fail on GauGAN images, since the generator structure of GauGAN (5 nearest neighbor up-samplers) is drastically different from the CycleGAN structure.”
BackgroundMethodInfluential
Cited by
Sorted by citations
Analyzing and Improving the Image Quality of StyleGAN
Karras, T., Laine, S., Aittala, M., Hellsten, J., Lehtinen, J., & Aila, T. (2020) · 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr42600.2020.00813 · Cited by 6,332
Deepfakes and beyond: A Survey of face manipulation and fake detection
Tolosana, R., Vera-Rodriguez, R., Fierrez, J., Morales, A., & Ortega-Garcia, J. (2020) · Information Fusion · doi:10.1016/j.inffus.2020.06.014 · Cited by 1,237
CNN-Generated Images Are Surprisingly Easy to Spot… for Now
Wang, S.-Y., Wang, O., Zhang, R., Owens, A., & Efros, A. A. (2020) · 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr42600.2020.00872 · Cited by 1,210
Media Forensics and DeepFakes: An Overview
Verdoliva, L. (2020) · IEEE Journal of Selected Topics in Signal Processing · doi:10.1109/jstsp.2020.3002101 · Cited by 979
“In [174], instead, a GAN simulator is proposed to reproduce common GAN-image artifacts, which manifest as spectral peaks in the Fourier domain.”
BackgroundThinking in Frequency: Face Forgery Detection by Mining Frequency-Aware Clues
Qian, Y., Yin, G., Sheng, L., Chen, Z., & Shao, J. (2020) · Lecture Notes in Computer Science · doi:10.1007/978-3-030-58610-2_6 · Cited by 1,058
Fake news, disinformation and misinformation in social media: a review
Aïmeur, E., Amri, S., & Brassard, G. (2023) · Social Network Analysis and Mining · doi:10.1007/s13278-023-01028-5 · Cited by 772
“…layers in neural networks place data in a hierarchy of different concepts, which ultimately learn from their own mistakes (Parrish Kaliyar et al. 2020; Zhang et al. 2019a; Deepak and Chitturi 2020; Shu et al. 2018a; Wang et al. 2019c) prefer to combine traditional machine learning and deep…”
Deepfake Detection: A Systematic Literature Review
Rana, M. S., Nobi, M. N., Murali, B., & Sung, A. H. (2022) · IEEE Access · doi:10.1109/access.2022.3154404 · Cited by 573
Khan, A. A., Chaudhari, O., & Chandra, R. (2023) · Expert Systems with Applications · doi:10.1016/j.eswa.2023.122778 · Cited by 677
Lips Don't Lie: A Generalisable and Robust Approach to Face Forgery Detection
Haliassos, A., Vougioukas, K., Petridis, S., & Pantic, M. (2021) · 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr46437.2021.00500 · Cited by 537
“Another line of research relates to the observation that fully CNN-generated images may exhibit anomalous patterns in the frequency spectrum [19, 18, 63, 45].”
BackgroundMasood, M., Nawaz, M., Malik, K. M., Javed, A., Irtaza, A., & Malik, H. (2023) · Applied Intelligence · doi:10.1007/s10489-022-03766-z · Cited by 543
“keypoints computed from the frequency domain, instead of employing raw sample pixels, was introduced in [168].”
BackgroundMethodInfluentialFocal Frequency Loss for Image Reconstruction and Synthesis
Jiang, L., Dai, B., Wu, W., & Loy, C. C. (2021) · IEEE/CVF International Conference on Computer Vision (ICCV) · doi:10.1109/iccv48922.2021.01366 · Cited by 541
“Recent studies [59, 72, 22] in media forensics have shown some notable periodic patterns in the frequency spectra of manipulated images, which may be consistent with artifacts in the spatial domain.”
BackgroundTwo-Branch Recurrent Network for Isolating Deepfakes in Videos
Masi, I., Killekar, A., Mascarenhas, R. M., Gurudatt, S. P., & AbdAlmageed, W. (2020) · Lecture Notes in Computer Science · doi:10.1007/978-3-030-58571-6_39 · Cited by 471
Towards Universal Fake Image Detectors that Generalize Across Generative Models
Ojha, U., Li, Y., & Lee, Y. J. (2023) · 2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr52729.2023.02345 · Cited by 428
“Earlier methods studied whether one can even learn a classifier that can detect other images from the same generative model [25, 33, 46], and later work found that such classifiers do not generalize to detecting fakes from other models [19, 52].”
BackgroundMethodInfluentialGlobal Texture Enhancement for Fake Face Detection in the Wild
Liu, Z., Qi, X., & Torr, P. H. S. (2020) · 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr42600.2020.00808 · Cited by 401
DeepFake Detection by Analyzing Convolutional Traces
Guarnera, L., Giudice, O., & Battiato, S. (2020) · 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW) · doi:10.1109/cvprw50498.2020.00341 · Cited by 374
“[44] proposed a method to classify Deepfakes considering the spectra of the frequency domain as input.”
MethodWhat Makes Fake Images Detectable? Understanding Properties that Generalize
Chai, L., Bau, D., Lim, S.-N., & Isola, P. (2020) · Lecture Notes in Computer Science · doi:10.1007/978-3-030-58574-7_7 · Cited by 357
Local Relation Learning for Face Forgery Detection
Chen, S., Yao, T., Chen, Y., Ding, S., Li, J., & Ji, R. (2021) · Proceedings of the AAAI Conference on Artificial Intelligence · doi:10.1609/aaai.v35i2.16193 · Cited by 334
DIRE for Diffusion-Generated Image Detection
Wang, Z., Bao, J., Zhou, W., Wang, W., Hu, H., Chen, H., & Li, H. (2023) · 2023 IEEE/CVF International Conference on Computer Vision (ICCV) · doi:10.1109/iccv51070.2023.02051 · Cited by 354
“[51] propose a GAN simulator, AutoGAN, to simulate the artifacts produced by standard GAN pipelines.”
MethodExploring Temporal Coherence for More General Video Face Forgery Detection
Zheng, Y., Bao, J., Chen, D., Zeng, M., & Wen, F. (2021) · 2021 IEEE/CVF International Conference on Computer Vision (ICCV) · doi:10.1109/iccv48922.2021.01477 · Cited by 320
“frequency, color) [42, 30, 19, 58, 20] or high-level semantics(e.”
On The Detection of Synthetic Images Generated by Diffusion Models
Corvi, R., Cozzolino, D., Zingarini, G., Poggi, G., Nagano, K., & Verdoliva, L. (2023) · ICASSP 2023 - 2023 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) · doi:10.1109/icassp49357.2023.10095167 · Cited by 304
“We compare the following detectors: Spec [9], based on frequency analysis, PatchForensics [18], that relies on local patch analysis, Wang2020 [15], a ResNet50 with blurring and compression augmentation, and Grag2021 [16], same backbone as before but avoiding down-sampling in the first layer and intense augmentation.”
BackgroundMethodRepresentative Forgery Mining for Fake Face Detection
Wang, C., & Deng, W. (2021) · 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr46437.2021.01468 · Cited by 288
“Recent remarkable breakthroughs [6, 8, 11, 13, 22, 40, 36, 38, 24] address this problem to some extent, which generally follow three directions: a) Through extracting the digital fin-gerprints produced by the defect of manipulation technique, some works [24, 36, 38, 40] achieve advanced…”
InfluentialFakeSpotter: A Simple yet Robust Baseline for Spotting AI-Synthesized Fake Faces
Wang, R., Juefei-Xu, F., Ma, L., Xie, X., Huang, Y., Wang, J., & Liu, Y. (2020) · International Joint Conference on Artificial Intelligence · doi:10.24963/ijcai.2020/476 · Cited by 313
“FakeSpotter also outperforms prior work AutoGAN [Zhang et al. , 2019b] and gives an average detection accuracy of more than 90% on the four types of fake faces.”
BackgroundResultInfluentialIMD2020: A Large-Scale Annotated Dataset Tailored for Detecting Manipulated Images
Novozámský, A., Mahdian, B., & Saic, S. (2020) · IEEE/CVF Winter Conference on Applications of Computer Vision Workshops (WACVW) · doi:10.1109/wacvw50321.2020.9096940 · Cited by 270
Identification of deep network generated images using disparities in color components
Li, H., Li, B., Tan, S., & Huang, J. (2020) · Signal Processing · doi:10.1016/j.sigpro.2020.107616 · Cited by 284
“For example, training the network with co-occurrence matrices extracted from an image in the pixel domain of RGB space [49], prompting the network to learning the affine face warping artifacts [50] or up-sampling artifacts [51].”
BackgroundHDNet: High-resolution Dual-domain Learning for Spectral Compressive Imaging
Hu, X., Cai, Y., Lin, J. F., Wang, H., Yuan, X., Zhang, Y., Timofte, R., & Gool, L. V. (2022) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr52688.2022.01702 · Cited by 243
“Recent studies [12,41,51] indicate that the periodic pattern shown in the frequency spectrum may be consistent with the artifacts in the spatial domain.”
BackgroundThe Stable Signature: Rooting Watermarks in Latent Diffusion Models
Fernandez, P., Couairon, G., Jégou, H., Douze, M., & Furon, T. (2023) · 2023 IEEE/CVF International Conference on Computer Vision (ICCV) · doi:10.1109/iccv51070.2023.02053 · Cited by 251
“Other approaches use traces left by the generators in the spatial [54, 97] or frequency [26, 106] domains.”
BackgroundMethodGenerative Adversarial Networks for Image and Video Synthesis: Algorithms and Applications
Liu, M.-Y., Huang, X., Yu, J., Wang, T.-C., & Mallya, A. (2021) · Proceedings of the IEEE · doi:10.1109/jproc.2021.3049196 · Cited by 236
“Many other approaches have been proposed to detect GAN-generated images across different generator architectures [204], [233].”
BackgroundDetecting Deep-Fake Videos from Appearance and Behavior
Agarwal, S., Farid, H., El-Gaaly, T., & Lim, S.-N. (2020) · 2020 IEEE International Workshop on Information Forensics and Security (WIFS) · doi:10.1109/wifs49906.2020.9360904 · Cited by 242
“The authors in [37] and [38] showed that GAN-generated content contains distinct digital fingerprints which can be learned and used to classify images as GAN-generated or not.”
BackgroundDetecting Deep-Fake Videos from Phoneme-Viseme Mismatches
Agarwal, S., Farid, H., Fried, O., & Agrawala, M. (2020) · 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW) · doi:10.1109/cvprw50498.2020.00338 · Cited by 234
“Some of these techniques detect generic artifacts [32, 30, 31, 28], while others detect explicit artifacts that result from, for example, image warping [20], image blending [18] and inconsistencies between the image and metadata [14].”
BackgroundHierarchical Fine-Grained Image Forgery Detection and Localization
Guo, X., Liu, X., Ren, Z., Grosz, S. A., Masi, I., & Liu, X. (2023) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr52729.2023.00308 · Cited by 251
“[78] report that it is difficult for classifiers to generalize across different GANs”
BackgroundMethodInfluentialDeepfake video detection: challenges and opportunities
Kaur, A., Noori Hoshyar, A., Saikrishna, V., Firmin, S., & Xia, F. (2024) · Artificial Intelligence Review · doi:10.1007/s10462-024-10810-6 · Cited by 214
Are GAN Generated Images Easy to Detect? A Critical Analysis of the State-Of-The-Art
Gragnaniello, D., Cozzolino, D., Marra, F., Poggi, G., & Verdoliva, L. (2021) · 2021 IEEE International Conference on Multimedia and Expo (ICME) · doi:10.1109/icme51207.2021.9428429 · Cited by 212
“In our analysis we compare a number of detectors: Xception [8], SRNet [28], Spec [9], M-Gb [15], Co-Net [6], Wang2020 [16], PatchForensics [17].”
Countering Malicious DeepFakes: Survey, Battleground, and Horizon
Juefei-Xu, F., Wang, R., Huang, Y., Guo, Q., Ma, L., & Liu, Y. (2022) · International Journal of Computer Vision · doi:10.1007/s11263-022-01606-8 · Cited by 191
“Figure 7 visualizes the spectra maps of real and fake face (Zhang et al., 2019), respectively.”
BackgroundInfluentialArtificial Fingerprinting for Generative Models: Rooting Deepfake Attribution in Training Data
Yu, N., Skripniuk, V., Abdelnabi, S., & Fritz, M. (2021) · 2021 IEEE/CVF International Conference on Computer Vision (ICCV) · doi:10.1109/iccv48922.2021.01418 · Cited by 195
“Most of them rely on low-level visual patterns in GAN-generated images [35, 53, 50, 20, 61] or frequency mismatch [14, 60, 15].”
BackgroundMethodDeepfakes Detection with Automatic Face Weighting
Montserrat, D. M., Hao, H., Yarlagadda, S. K., Baireddy, S., Shao, R., Horvath, J., Bartusiak, E., Yang, J., Guera, D., Zhu, F., et al. (2020) · 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW) · doi:10.1109/cvprw50498.2020.00342 · Cited by 181
“In [17, 18, 19], forensic analysis of GAN generated images revealed that GANs leave some high frequency fingerprints in the images they generate.”
DE-FAKE: Detection and Attribution of Fake Images Generated by Text-to-Image Generation Models
Sha, Z., Li, Z., Yu, N., & Zhang, Y. (2023) · Proceedings of the 2023 ACM SIGSAC Conference on Computer and Communications Security · doi:10.1145/3576915.3616588 · Cited by 175
“[39], we draw the frequency spectra of fake and real images.”
Evading Deepfake-Image Detectors with White- and Black-Box Attacks
Carlini, N., & Farid, H. (2020) · 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW) · doi:10.1109/cvprw50498.2020.00337 · Cited by 170
“Low-level forensic techniques detect pixel-level artifacts introduced by the synthesis process [47, 34, 39, 48].”
BackgroundDeep fake geography? When geospatial data encounter Artificial Intelligence
Zhao, B., Zhang, S., Xu, C., Sun, Y., & Deng, C. (2021) · Cartography and Geographic Information Science · doi:10.1080/15230406.2021.1910075 · Cited by 181
“Scholars found that GAN-generated fake images were different from authentic ones in multiple visual features such as color, texture and details, and in frequency domain features such as a certain type of periodic replications (Galbally & Marcel, 2014; Wang et al., 2020; X. Zhang et al., 2019).”
BackgroundResultFighting Deepfake by Exposing the Convolutional Traces on Images
Guarnera, L., Giudice, O., & Battiato, S. (2020) · IEEE Access · doi:10.1109/access.2020.3023037 · Cited by 149
“[17] (the authors of AutoGAN) and Wang et al.”
BackgroundInfluentialExposing GAN-Generated Faces Using Inconsistent Corneal Specular Highlights
Hu, S., Li, Y., & Lyu, S. (2021) · IEEE International Conference on Acoustics Speech and Signal Processing · doi:10.1109/icassp39728.2021.9414582 · Cited by 150
“More recent works such as [17, 18, 19] extend the analysis to the frequency domain, where the upsampling step in the GAN generation leaves specific artifacts.”
BackgroundFrePGAN: Robust Deepfake Detection Using Frequency-Level Perturbations
Jeong, Y., Kim, D., Ro, Y., & Choi, J. (2022) · Proceedings of the AAAI Conference on Artificial Intelligence · doi:10.1609/aaai.v36i1.19990 · Cited by 137
Combating disinformation in a social media age
Shu, K., Bhattacharjee, A., Alatawi, F., Nazer, T. H., Ding, K., Karami, M., & Liu, H. (2020) · WIREs Data Mining and Knowledge Discovery · doi:10.1002/widm.1385 · Cited by 198
“Utilizing these artifacts, the authors in Zhang et al. (2019) have conducted experiments to classify real versus fake GAN generated images, achieving varying degrees of success for images from different domains.”
BackgroundA Survey on the Detection and Impacts of Deepfakes in Visual, Audio, and Textual Formats
Mubarak, R., Alsboui, T., Alshaikh, O., Inuwa-Dutse, I., Khan, S., & Parkinson, S. (2023) · IEEE Access · doi:10.1109/access.2023.3344653 · Cited by 132
“Zhang et al. [162] also focused on identifying and simulating artefacts unique to GAN-generated images using frequency spectrum features; however, their model was limited to image detection only.”
BackgroundIntriguing properties of synthetic images: from generative adversarial networks to diffusion models
Corvi, R., Cozzolino, D., Poggi, G., Nagano, K., & Verdoliva, L. (2023) · 2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW) · doi:10.1109/cvprw59228.2023.00104 · Cited by 142
“More specifically, in [59] for the first time evidence of such traces was shown, which appear as peaks in the Fourier domain and can be used as discriminating features for the development of a forensic detector.”
Raising the Bar of AI-generated Image Detection with CLIP
Cozzolino, D., Poggi, G., Corvi, R., Nießner, M., & Verdoliva, L. (2024) · IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW) · doi:10.1109/cvprw63382.2024.00439 · Cited by 181
“This phenomenon was first observed for methods based on generative adversarial networks (GAN) [20, 73] but holds, with obvious differences, for all generation approaches, including the more recent diffusion models (DM) [11].”
InfluentialGenimage: A million-scale benchmark for detecting ai-generated image
Zhu, M., Chen, H., Yan, Q., Huang, X., Lin, G., Li, W., Tu, Z., Hu, H., Hu, J., & Wang, Y. (2023) · Advances in Neural Information Processing Systems 36 · doi:10.52202/075280-3398 · Cited by 128
Fighting Deepfakes by Detecting GAN DCT Anomalies
Giudice, O., Guarnera, L., & Battiato, S. (2021) · Journal of Imaging · doi:10.3390/jimaging7080128 · Cited by 118
“In [3], it has been already demonstrated that it is possible to attack and retrieve the signature on the network’s de-convolutional layers; in this paper a method to identify any anomaly of the generated “fake” signal, only partially highlighted in some preliminary studies [4], [5] is presented.”
InfluentialA detailed analysis of image and video forgery detection techniques
Tyagi, S., & Yadav, D. (2023) · The Visual Computer · doi:10.1007/s00371-021-02347-4 · Cited by 109
“[84] (2019) Spectrum domain features GAN discriminaor Created by authors from (StarGAN/CycleGAN) Acc.”
BackgroundAI-enabled image fraud in scientific publications
Gu, J., Wang, X., Li, C., Zhao, J., Fu, W., Liang, G., & Qiu, J. (2022) · Patterns · doi:10.1016/j.patter.2022.100511 · Cited by 103
Face Forgery Detection by 3D Decomposition
Zhu, X., Wang, H., Fei, H., Lei, Z., & Li, S. Z. (2021) · 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr46437.2021.00295 · Cited by 106
“[41] identify the”
BackgroundBiHPF: Bilateral High-Pass Filters for Robust Deepfake Detection
Jeong, Y., Kim, D., Min, S., Joe, S., Gwon, Y., & Choi, J. (2022) · 2022 IEEE/CVF Winter Conference on Applications of Computer Vision (WACV) · doi:10.1109/wacv51458.2022.00293 · Cited by 106
FakeLocator: Robust Localization of GAN-Based Face Manipulations
Huang, Y., Juefei-Xu, F., Guo, Q., Liu, Y., & Pu, G. (2022) · IEEE Transactions on Information Forensics and Security · doi:10.1109/tifs.2022.3141262 · Cited by 102
“[36] has proved that the transpose convolution and nearest neighbor interpolation have fake texture.”
BackgroundSynthbuster: Towards Detection of Diffusion Model Generated Images
Bammey, Q. (2023) · IEEE Open Journal of Signal Processing · doi:10.1109/ojsp.2023.3337714 · Cited by 125
Kim, B., Xiong, A., Lee, D., & Han, K. (2021) · PLoS ONE · doi:10.1371/journal.pone.0260080 · Cited by 113
A Closer Look at Fourier Spectrum Discrepancies for CNN-generated Images Detection
Chandrasegaran, K., Tran, N.-T., & Cheung, N.-M. (2021) · 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr46437.2021.00712 · Cited by 90
“worked reasonably with RGB inputs [42, 5], with spectral discrepancies being observed, several works have proposed to use the corresponding Fourier representation to train these detectors [44, 13].”
BackgroundMethodDetecting fake images by identifying potential texture difference
Yang, J., Xiao, S., Li, A., Lan, G., & Wang, H. (2021) · Future Generation Computer Systems · doi:10.1016/j.future.2021.06.043 · Cited by 86
Detecting DeepFake, FaceSwap and Face2Face facial forgeries using frequency CNN
Kohli, A., & Gupta, A. (2021) · Multimedia Tools and Applications · doi:10.1007/s11042-020-10420-8 · Cited by 90
“The proposed fCNN has a simple 3 convolutional layered architecture as compared to GAN based complex architectures in [2] and [31].”
BackgroundMethodInfluentialDetecting Generated Images by Real Images
Liu, B., Yang, F., Bi, X., Xiao, B., Li, W., & Gao, X. (2022) · Lecture notes in computer science · doi:10.1007/978-3-031-19781-9_6 · Cited by 86
“Zhang et al. [15] introduce a generator that simulates sampling artifacts on several common GANs and demonstrates superior performance in the frequency domain by learning to classify sampling artifacts on GANs in both the spatial and frequency domains.”
BackgroundMethodWang, R., Juefei-Xu, F., Luo, M., Liu, Y., & Wang, L. (2021) · ACM International Conference on Multimedia (ACM MM) · doi:10.1145/3474085.3475518 · Cited by 97
“In the black-box setting, we employ a GAN simulator proposed in AutoGAN [65] to simulate the generation of DeepFake transformation.”
BackgroundMethodInfluentialEvading DeepFake Detectors via Adversarial Statistical Consistency
Hou, Y., Guo, Q., Huang, Y., Xie, X., Ma, L., & Zhao, J. (2023) · 2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr52729.2023.01181 · Cited by 76
“For the frequency-based detectors, we adopt two popular detectors, DCTA [10] and DFTD [53].”
BackgroundMethodInfluentialMastering Deepfake Detection: A Cutting-Edge Approach to Distinguish GAN and Diffusion-Model Images
Guarnera, L., Giudice, O., & Battiato, S. (2024) · ACM Transactions on Multimedia Computing, Communications, and Applications · doi:10.1145/3652027 · Cited by 74
“The analysis in the Fourier domain was employed by Zhang et al. [46] with AutoGAN in a rather naive strategy that delivered, in any case, good performances.”
InfluentialThink Twice Before Detecting GAN-generated Fake Images from their Spectral Domain Imprints
Dong, C., Kumar, A., & Liu, E. (2022) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr52688.2022.00771 · Cited by 72
“[45] propose a spectrum detector with good generalizing ability which is trained on spectra of images synthesized by a GAN simulator that shares the architectures which are common in most GANs.”
BackgroundMethodProactive Deepfake Defence via Identity Watermarking
Zhao, Y., Liu, B., Ding, M., Liu, B., Zhu, T., & Yu, X. (2023) · IEEE Winter Conference on Applications of Computer Vision · doi:10.1109/wacv56688.2023.00458 · Cited by 76
“AutoGAN [53] first observes that the up-sampling design in GAN would introduce artifacts in the synthesized images which can be used to detect GAN-generated images.”
BackgroundDetecting Deep-Fake Videos from Aural and Oral Dynamics
Agarwal, S., & Farid, H. (2021) · IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW) · doi:10.1109/cvprw53098.2021.00109 · Cited by 70
“Low-level techniques detect pixel-level, synthesis artifacts, including generic artifacts [41, 37, 40, 35], warping artifacts [26], and blending artifacts [24].”
BackgroundWavelet-packets for deepfake image analysis and detection
Wolter, M., Blanke, F., Heese, R., & Garcke, J. (2022) · Machine Learning · doi:10.1007/s10994-022-06225-5 · Cited by 84
“…deep network generated images in the frequency space created by the discrete cosine transform (DCT), as well as detectors based on Fourier features (Zhang et al., 2019; Durall et al., 2019; Dzanic et al., 2020; Frank et al., 2020; Durall et al., 2020; Giudice et al., 2021), these provide frequency…”
BackgroundBeyond the Spectrum: Detecting Deepfakes via Re-Synthesis
He, Y., Yu, N., Keuper, M., & Fritz, M. (2021) · Proceedings of the Thirtieth International Joint Conference on Artificial Intelligence · doi:10.24963/ijcai.2021/349 · Cited by 70
Supervised Contrastive Learning for Generalizable and Explainable DeepFakes Detection
Xu, Y., Raja, K., & Pedersen, M. (2022) · 2022 IEEE/CVF Winter Conference on Applications of Computer Vision Workshops (WACVW) · doi:10.1109/wacvw54805.2022.00044 · Cited by 62
AI vs. AI: Can AI Detect AI-Generated Images?
Baraheem, S. S., & Nguyen, T. (2023) · Journal of Imaging · doi:10.3390/jimaging9100199 · Cited by 79
“Other studies Other studies have worked on the Fourier domain to detect GAN-generated images from real ones [11–13]. have worked on the Fourier domain to detect GAN-generated images from real ones [11–13].”
BackgroundDeepfakes: A new threat to image fabrication in scientific publications?
Wang, L., Zhou, L., Yang, W., & Yu, R. (2022) · Patterns · doi:10.1016/j.patter.2022.100509 · Cited by 60
“8 A well-trained generator may gloss over this effect, making it hard to be discovered by the naked eye in the spatial domain.”
BackgroundReverse Engineering of Generative Models: Inferring Model Hyperparameters From Generated Images
Asnani, V., Yin, X. P., Hassner, T., & Liu, X. (2023) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2023.3301451 · Cited by 73
“Motivatedbypriorworksonphysicalfingerprintestimation[17], [22],[23],[24],[48],wedefinethefollowingfourconstraintsto guideourestimatedfingerprintstohavethedesirableproperties.”
BackgroundBinh, L. M., & Woo, S. (2022) · Proceedings of the AAAI Conference on Artificial Intelligence · doi:10.1609/aaai.v36i1.19886 · Cited by 63
An Exploratory Analysis on Visual Counterfeits Using Conv-LSTM Hybrid Architecture
Hashmi, M. F., Ashish, B. K. K., Keskar, A. G., Bokde, N. D., Yoon, J. H., & Geem, Z. W. (2020) · IEEE Access · doi:10.1109/access.2020.2998330 · Cited by 62
“Zhang et al. [37] find that classifiers generalize poorly between GAN models.”
BackgroundDeepfake Network Architecture Attribution
Yang, T., Huang, Z., Cao, J., Li, L., & Li, X. (2022) · Proceedings of the AAAI Conference on Artificial Intelligence · doi:10.1609/aaai.v36i4.20391 · Cited by 62
Making DeepFakes More Spurious: Evading Deep Face Forgery Detection via Trace Removal Attack
Liu, C., Chen, H., Zhu, T., Zhang, J., & Zhou, W. (2023) · IEEE Transactions on Dependable and Secure Computing · doi:10.1109/tdsc.2023.3241604 · Cited by 69
“Zhang et al. [37] identify unknown DeepFakes by detecting and simulating general spectrum artifacts.”
BackgroundGLFF: Global and Local Feature Fusion for AI-Synthesized Image Detection
Ju, Y., Jia, S., Cai, J., Guan, H., & Lyu, S. (2023) · IEEE Transactions on Multimedia · doi:10.1109/tmm.2023.3313503 · Cited by 62
“Representatives such as spectrum magnitude, 2d-FFT, and 2d-DCT serve as input features for CNN binary classification [17], [47], [51], [52].”
BackgroundMethodTowards Discovery and Attribution of Open-world GAN Generated Images
Girish, S., Suri, S., Rambhatla, S. S., & Shrivastava, A. (2021) · IEEE/CVF International Conference on Computer Vision (ICCV) · doi:10.1109/iccv48922.2021.01383 · Cited by 61
“Therefore, a plethora of works have been proposed which focus on the real/fake detection task [4, 5, 6].”
BackgroundIdentifying and Mitigating the Security Risks of Generative AI
Barrett, C., Boyd, B., Bursztein, E., Carlini, N., Chen, B., Choi, J., Chowdhury, A. R., Christodorescu, M., Datta, A., Feizi, S., et al. (2023) · Foundations and Trends® in Privacy and Security · doi:10.1561/3300000041 · Cited by 57
“Zero-shot detectors: These detectors perform without any additional training overhead and use some statistical signatures of AI-generated content to conduct the detection [84, 39, 32].”
MethodLaRE 2 : Latent Reconstruction Error Based Method for Diffusion-Generated Image Detection
Luo, Y., Du, J., Yan, K., & Ding, S. (2024) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr52733.2024.01609 · Cited by 68
“Some works [10, 44] also find that there are obvious visual artifacts in the GAN-generated images and detect these images from the frequency view [28].”
BackgroundFakePolisher: Making DeepFakes More Detection-Evasive by Shallow Reconstruction
Huang, Y., Juefei-Xu, F., Wang, R., Guo, Q., Ma, L., Xie, X., Li, J., Miao, W., Liu, Y., & Pu, G. (2020) · Proceedings of the 28th ACM International Conference on Multimedia · doi:10.1145/3394171.3413732 · Cited by 71
“As [59] mentioned, the GAN-synthesised im-”
BackgroundSpoC: Spoofing Camera Fingerprints
Cozzolino, D., Thies, J., Rossler, A., Niesner, M., & Verdoliva, L. (2021) · 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW) · doi:10.1109/cvprw53098.2021.00110 · Cited by 54
“In addition, we consider the spectrum-based (Spec) classifier4 proposed in [59], which detects the frequency-domain peaks caused by the up-sampling steps of common GAN pipelines.”
BackgroundDeepfake attribution: On the source identification of artificially generated images
Khoo, B., Phan, R. C. ‐., & Lim, C. H. (2021) · Wiley Interdisciplinary Reviews Data Mining and Knowledge Discovery · doi:10.1002/widm.1438 · Cited by 62
“In 2016 when GAN models were still restricted to low-resolution images, Odena et al. (2016) proved the existence of “ checkerboard artifacts ” in all images upsampled by ConvNets.”
BackgroundMethodInfluentialRobust Attentive Deep Neural Network for Detecting GAN-Generated Faces
Hui, G., Hu, S., Wang, X., Chang, M., & Lyu, S. (2022) · IEEE Access · doi:10.1109/access.2022.3157297 · Cited by 62
“Many works [36]–[38] identify GAN images by recognizing the specific artifacts produced by the GAN upsampling process.”
BackgroundDynamic texture analysis for detecting fake faces in video sequences
Bonomi, M., Pasquini, C., & Boato, G. (2021) · Journal of Visual Communication and Image Representation · doi:10.1016/j.jvcir.2021.103239 · Cited by 61
Li, J., Chen, L., Jian, S., Xiao, X., Liu, X., Sun, X., Wang, X., & Li, D. (2023) · Remote Sensing · doi:10.3390/rs15020440 · Cited by 51
NTIRE 2020 Challenge on Perceptual Extreme Super-Resolution: Methods and Results
Zhang, K., Gu, S., Timofte, R., Shang, T., Dai, Q., Zhu, S., Yang, T., Guo, Y., Jo, Y., Yang, S., et al. (2020) · IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW) · doi:10.1109/cvprw50498.2020.00254 · Cited by 51
SSD-GAN: Measuring the Realness in the Spatial and Spectral Domains
Chen, Y., Li, G., Jin, C., Liu, S., & Li, T. (2021) · Proceedings of the AAAI Conference on Artificial Intelligence · doi:10.1609/aaai.v35i2.16196 · Cited by 51
MCS-GAN: A Different Understanding for Generalization of Deep Forgery Detection
Xiao, S., Lan, G., Yang, J., Lu, W. B., Meng, Q., & Gao, X. (2023) · IEEE Transactions on Multimedia · doi:10.1109/tmm.2023.3279993 · Cited by 51
On the Frequency Bias of Generative ModelsPreprint
Schwarz, K., Liao, Y., & Geiger, A. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2111.02447 · Cited by 50
“While GAN training is notoriously hard, learning the distribution of high-frequency content is particularly problematic [8– 10, 15, 24, 37, 41].”
BackgroundInfluentialShen, Z., Ding, F., Jolfaei, A., Yadav, K., Vashisht, S., & Yu, K. (2022) · IEEE Transactions on Network Science and Engineering · doi:10.1109/tnse.2022.3190765 · Cited by 49
“From their experiment set-up [41], we noticed that there was another point to be mentioned.”
BackgroundMethodGao, J., Micheletto, M., Orrù, G., Concas, S., Feng, X., Marcialis, G. L., & Roli, F. (2024) · Engineering Applications of Artificial Intelligence · doi:10.1016/j.engappai.2024.108450 · Cited by 56
Ensembling with Deep Generative Views
Chai, L., Zhu, J.-Y., Shechtman, E., Isola, P., & Zhang, R. (2021) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr46437.2021.01475 · Cited by 50
“that despite their rapid improvement, the generative models do, in fact, still exhibit artifacts, exploitable for reliable detection of GAN-generated imagery [59, 10, 62, 66, 16].”
Proactive Image Manipulation Detection
Asnani, V., Yin, X. P., Hassner, T., Liu, S., & Liu, X. (2022) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr52688.2022.01495 · Cited by 48
“We show that our framework achieves superior manipulation detection than State-of-TheArt (SoTA) methods [10, 28, 46, 59].”
BackgroundMethodResultInfluentialFFR_FD: Effective and fast detection of DeepFakes via feature point defects
Wang, G., Jiang, Q., Jin, X., & Cui, X. (2022) · Information Sciences · doi:10.1016/j.ins.2022.03.026 · Cited by 50
“The usampling strategies of deep generative models introduce artifacts in the frequency domain [68, 69], inspiring many spectrum-based detection methods [22, 70].”
Cazenavette, G., Sud, A., Leung, T., & Usman, B. (2024) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr52733.2024.01023 · Cited by 59
“…patches, Ju et al. [27] explored fusing global and local image features, Corvi et al. [17] explored better augmentation and downsampling strategies, Zhang et al. [61] and Frank et al. [20] explored artifacts in the spectrum of GAN-generated images, and Marra et al. [35] explored GAN fingerprinting.”
BackgroundDetecting Deepfake Videos using Attribution-Based Confidence Metric
Fernandes, S. L., Raj, S., Ewetz, R., Pannu, J. S., Jha, S. K., Ortiz, E., Vintila, I., & Salter, M. (2020) · IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW) · doi:10.1109/cvprw50498.2020.00162 · Cited by 47
“Zhan et al [59], performed an analysis on the spectrum domain instead of considering raw pixel information.”
BackgroundWang, G., Jiang, Q., Jin, X., Li, W., & Cui, X. (2022) · Knowledge-Based Systems · doi:10.1016/j.knosys.2022.109114 · Cited by 49
C2P-CLIP: Injecting Category Common Prompt in CLIP to Enhance Generalization in Deepfake Detection
Tan, C., Tao, R., Liu, H., Gu, G., Wu, B., Zhao, Y., & Wei, Y. (2025) · Proceedings of the AAAI Conference on Artificial Intelligence · doi:10.1609/aaai.v39i7.32772 · Cited by 93
Detecting Real-Time Deep-Fake Videos Using Active Illumination
Gerstner, C. R., & Farid, H. (2022) · IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW) · doi:10.1109/cvprw56347.2022.00015 · Cited by 45
Face Forgery Detection by 3D Decomposition and Composition Search
Zhu, X., Fei, H., Zhang, B., Zhang, T., Zhang, X., Li, S. Z., & Lei, Z. (2023) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2022.3233586 · Cited by 41
“…image forgery detection [24], [25], [26], and recent advances in DL promote the learning-based methods [27], [28], [29], [30], [31], not only revealing inter-domain artifacts of the generative model [12], [32], [33], but also sub-stantiating the effectiveness of simple supervised classifiers [34].”
BackgroundSelf-citationZhou, Y., Koyuncu, C., Lu, C. Y., Grobholz, R., Katz, I. A., Madabhushi, A., & Janowczyk, A. R. (2022) · Medical Image Analysis · doi:10.1016/j.media.2022.102702 · Cited by 40
“For instance, GANs are known to introduce artifacts due to the loss of high-frequency content, and are also capable of learning histologic features other than pre-analytic variations (e.g., stain) from templates (Wei et al., 2019; Zhang et al., 2019).”
Background
Metadata
- Authors
- Zhang, X., Karaman, S., & Chang, S.-F.
- Affiliations
- Columbia University (US)
- Venue
- 2019 IEEE International Workshop on Information Forensics and Security (WIFS)
- Publisher
- IEEE
- Published
- 2019-12
- Pages
- 1-6
- ISSN
- 2157-4774 (ISSN-L)
- OpenAlex
- W3012472557
- Cited by
- 567
- References
- 21
- Citations by year
- 2026: 60 · 2025: 113 · 2024: 89 · 2023: 100 · 2022: 84 · 2021: 81 · 2020: 39
- FWCI
- 16.52
- Type
- conference paper
- Language
- en
- Open access
- green
- arxiv.org
- Licence
- Locations
- IEEE International Workshop on Information Forensics and Security (publishedVersion)
- arXiv (Cornell University) (submittedVersion)
- arXiv (Cornell University)
- Topics
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Digital Media Forensic Detection
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Generative Adversarial Networks and Image Synthesis
- Physical Sciences › Computer Science › Artificial Intelligence › Adversarial Robustness in Machine Learning
- Keywords
- GAN-generated image detection, CycleGAN, adversarial image generation
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
