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
Items 401–500 of 561
Combating Dataset Misalignment for Robust AI-Generated Image Detection in the Real World
Choi, H., Jung, I., & Woo, S. S. (2025) · doi:10.1145/3709022.3736541 · Cited by 3
“The ForenSynths [35] dataset is one of the most widely used benchmarks in generated image detection [3, 4, 22, 24, 26, 32, 34, 35, 39], benefiting from its large-scale and diverse set of both real and fake images.”
DeepFake Detection Using CNN Models with Explainable AI Techniques
Bipin, P. R., Biju, A. K., Balagopal, M., Shyam, B., & Adithya, C. (2025) · International Conference on Advances in Computing, Communication, Embedded and Secure Systems (ACCESS) · doi:10.1109/access65134.2025.11135663 · Cited by 3
“Zhang et al. [20] explored GAN-generated image artifacts using a transformer-based approach.”
mToFNet: Object Anti-Spoofing with Mobile Time-of-Flight Data
Jeong, Y., Kim, D., Lee, J., Hong, M., Hwang, S., & Choi, J. (2022) · IEEE Winter Conference on Applications of Computer Vision · doi:10.1109/wacv51458.2022.00305 · Cited by 2
$$D^3$$: A Novel Face Forgery Detector Based on Dual-Stream and Dual-Utilization Methods
Wang, J., Li, X., & Zhao, Y. (2022) · Communications in computer and information science · doi:10.1007/978-3-031-06767-9_34 · Cited by 2
Survey of Fake Image Synthesis and its Detection
Uloli, T., Akash, R. M. K., Keerthika, A., & Dhanwanth, K. S. (2023) · Journal of Innovative Image Processing · doi:10.36548/jiip.2022.4.006 · Cited by 2
A review of image authentication techniques based on deep hashing
Lu, T., Feng, L., & Dong, Q. (2022) · International Conference on Computer Graphics, Image and Virtualization (ICCGIV) · doi:10.1109/iccgiv57403.2022.00021 · Cited by 2
Gamazo, J., Troncoso, J. M. C., & Rincón, M. (2022) · Lecture notes in computer science · doi:10.1007/978-3-031-06527-9_41 · Cited by 2
Are High-Frequency Components Beneficial for Training of Generative Adversarial Networks.Preprint
Li, Z., Xia, P., Xue, R., Hu, Y., & Li, B. (2021) · arXiv (Cornell University) · Cited by 2
Could Human Gaze Augment Detectors of Synthetic Images?
Fotopoulos, N., Riedmiller, C., Bozkir, E., Tsinganos, P., Ampeliotis, D., Kasneci, G., Kasneci, E., & Skodras, A. N. (2023) · IEEE International Conference on Digital Signal Processing (DSP) · doi:10.1109/dsp58604.2023.10167876 · Cited by 1
“[6] exploited the fact that the up-sampler of the GAN generator introduces image artifacts, which manifest as periodic peaks in the Fourier spectrum.”
BackgroundUnveiling Universal Forensics of Diffusion Models with Adversarial Perturbations
Xie, K., Liu, J., Zhu, M., Ding, G., Liu, Z., Chen, H., & Chen, H. (2024) · International Joint Conference on Neural Networks (IJCNN) · doi:10.1109/ijcnn60899.2024.10650754 · Cited by 1
“However, AI-synthesized content holds the power to entertain but also to deceive [5], [6].”
BackgroundKaman, S., & Makandar, A. (2024) · IEEE International Conference on Electronics, Computing and Communication Technologies (CONECCT) · doi:10.1109/conecct62155.2024.10677133 · Cited by 1
“Zhang et al. [7] investigated the GAN fingerprint [8] and suggested the Auto-GAN classifier model, which is based on the frequency spectrum input.”
MethodYour Diffusion Model is an Implicit Synthetic Image Detector
Wang, X., & Kalogeiton, V. S. (2025) · Lecture notes in computer science · doi:10.1007/978-3-031-92648-8_25 · Cited by 2
Hu, X., Haihong, E., Liu, J., Hu, T., Qiao, X., Ding, J., & Huang, J. (2024) · International Conference on Behavioral, Economic, and Socio-Cultural Computing · doi:10.1109/besc64747.2024.10780603 · Cited by 1
“Our goal is to demonstrate the effectiveness of our model on different datasets relative to the main evaluation metrics. a) Benchmarking methods Zhang et al. WIFS'19 [13]: This method utilizes spectral signatures for detection and has shown promising results in early tests.”
Multi-domain Multi-scale DeepFake Detection for Generalization
Liu, T., Liu, Y., Li, N., & Shi, H. (2024) · International Conference on Electronic Information Engineering and Computer Communication (EIECC) · doi:10.1109/eiecc64539.2024.10929576 · Cited by 1
“Research indicates that even low-quality images contain detectable frequency-domain artifacts, making them valuable forensic clues [15, 16] .”
Media Forensics and Deepfake-Systematic Survey
Nadeem Jabbar, C. H., Saghir, A., Meer, A. A., Sahi, S., Hassan, B., & Yasir, S. M. (2025) · Communications in computer and information science · doi:10.1007/978-3-031-77617-5_24 · Cited by 6
Fair-FLIP: Fair Deepfake Detection with Fairness-Oriented Final Layer Input Prioritising
Szandała, T., Ezzeddine, F., Rusin, N., Giordano, S., & Ayoub, O. (2025) · IEEE Swiss Conference on Data Science (SDS) · doi:10.1109/sds66131.2025.00021 · Cited by 3
“Several studies have focused on detecting deepfake across images, audio, and video modalities [2], [3], [6], [10]–[13], leveraging statistical techniques [30]–[33], ML-based [34]– [36] or deep learning-based methods [37]–[42].”
Building a Universal Detector of AI‐Generated Images Without Training on Them
Li, J., & Wang, K. (2025) · International Journal of Intelligent Systems · doi:10.1155/int/8530953 · Cited by 1
“Zhang et al. [23] proposed a GAN simulator to replicate artifacts arising from the shared operational pipelines of GAN architectures.”
Keïta, M., Hamidouche, W., Eutamene, H. B., Ahmed, A. T., & Hadid, A. (2025) · doi:10.1145/3746265.3759667 · Cited by 2
Rethinking Individual Fairness in Deepfake Detection
Hou, A., Lin, L., Li, J., & Hu, S. (2025) · ACM International Conference on Multimedia (ACM MM) · doi:10.1145/3746027.3755244 · Cited by 2
“Kearns et al. [60] proposed average individual fairness, which balances statistical and individual fairness by ensuring fairness across both individuals and classification tasks.”
Detecting AI-Generated Content
de Melo, G. (2025) · Oxford University Press eBooks · doi:10.1093/oxfordhb/9780198940272.013.0008 · Cited by 1
Advancing Deepfake Detection in RGB Satellite Imagery Through Domain-Specific Ensembles
Ziulu, V., & Garforth, J. (2025) · IEEE International Geoscience and Remote Sensing Symposium (IGARSS) · doi:10.1109/igarss55030.2025.11243512 · Cited by 1
“To evaluate robustness [44, 43, 54, 55, 59–63], models were tested on perturbed datasets where 50% of images were randomly subjected to (i) blurring: Gaussian blur with kernel size 3, 7, 11, 15 (chosen randomly), (ii) Gaussian Noise, (iii) salt and pepper noise (amount of noise in each image was…”
Adaptive Test-Time Semantic Debiasing for AI-Generated Image Detection
Cai, Y., Tian, J., Fu, X., Dai, J., Han, J., & Lyu, S. (2025) · IEEE International Conference on Computer Vision Workshops · doi:10.1109/iccvw69036.2025.00165 · Cited by 1
CatAID: Category-Guided AI-Generated Image Detection via Vision-Language Model Adaptation
Cai, Y., Jia, S., Tian, J., Dai, J., Han, J., & Lyu, S. (2025) · IEEE International Conference on Computer Vision Workshops · doi:10.1109/iccvw69036.2025.00166 · Cited by 1
“Signal-level methods primarily focus on the low-level patterns of images, such as noise or frequency artifacts left by the generation process [63, 68, 72].”
CLIP-ADA: CLIP-Guided Artifact-Invariant Generalizable Synthetic Image Detection
Deng, J., Zhang, C., Xu, C., Lin, C., Zhao, Z., Liu, S., Wang, Q., & Shen, C. Q. (2026) · IEEE Transactions on Information Forensics and Security · doi:10.1109/tifs.2026.3676657 · Cited by 1
“The frequency artifacts are validated to be introduced by the up-sampling operation of generative models [12], [33] and therefore several methods perform detection [13], [14], [15] based on it.”
Jahan, F., Shifat, S. M., Morol, M. K., Fahad, N., Goh, K. O. M., Nandi, D., Abdullah-Al-Jubair, M., & Hossen, M. J. (2026) · Discover Applied Sciences · doi:10.1007/s42452-026-08402-w · Cited by 2
LRD-Net: A Lightweight Real-Centered Detection Network for Cross-Domain Face Forgery Detection
Zhang, X., & Chaudhary, V. (2026) · International Symposium on Digital Forensics and Security (ISDFS) · doi:10.1109/isdfs69419.2026.11458948 · Cited by 1
“To address this limitation, we propose the Multi-Scale Wavelet Guidance Module (MSWGM) , which decomposes the input image into multiple frequency bands and learns to emphasize the most discriminative scales for forgery detection.”
Self-citationDiffusion Epistemic Uncertainty with Asymmetric Learning for Diffusion-Generated Image Detection
Huang, Y., Guo, H., Huang, J., Bai, B., & Xiong, Q. (2025) · IEEE/CVF International Conference on Computer Vision (ICCV) · doi:10.1109/iccv51701.2025.01588 · Cited by 3
“Accordingly, the community shifted to fine-grained spatial [4, 18, 52] and spectral signatures [16, 45, 54, 61], work that was initially developed for GAN imagery [33].”
DDSF-Net: dual-domain deepfake detection via semantic suppression and spatial-frequency mapping
Xing, Z., Liu, L., Wu, Y., Chang, C.-C., Wang, A., & Chang, C.-C. (2026) · Expert Systems with Applications · doi:10.1016/j.eswa.2026.133102 · Cited by 1
Using Grayscale Frequency Statistic to Detect Manipulated Faces in Wavelet-Domain
Wang, G., Li, W., Jiang, Q., Jin, X., & Cui, X. (2021) · Conference proceedings/Conference proceedings - IEEE International Conference on Systems, Man, and Cybernetics · doi:10.1109/smc52423.2021.9658971 · Cited by 1
“The recent attempts to finding the differences between real and fake images are in the frequency domain [10], [11], [12], [13], [36], and these methods are inspired by the up-sampling strategy of GANs when producing simulation data.”
MethodDetection of artificial fragments embedded in remote sensing images by adversarial neural networks
Gashnikov, M. V., & Кузнецов, А. В. (2022) · Computer Optics · doi:10.18287/2412-6179-co-1064 · Cited by 1
FAD-Net: Fake Images Detection and Generalization Based on Frequency Domain Transformation
Liu, X., Liu, J., Guo, P., Tuo, D., Tian, S., & Jiang, Y. (2022) · International Congress on Image and Signal Processing (CISP) · doi:10.1109/cisp-bmei56279.2022.9980271 · Cited by 1
“This paper mainly compares with Zhang et al.[7] and Wang et al.[12] This paper compares the model's generalization effect from two aspects: average precision (AP) and accuracy (ACC).”
MethodResultInfluentialGAN-Generated Face Detection with Strong Generalization Ability Based on Quaternions
Chen, B., Li, T., Wang, J., & Zhao, G. (2022) · Journal of Computer-Aided Design & Computer Graphics · doi:10.3724/sp.j.1089.2022.19015 · Cited by 1
“Zhang 等 [19] 也发现 , 与频域相比 , 空间域 更能凸显 GAN 生成图像的共性 .”
InfluentialImage Inpainting Anti-Forensics Network via Attention-Guided Hierarchical Reconstruction
Dou, L., Feng, G., & Qian, Z. (2023) · Symmetry · doi:10.3390/sym15020393 · Cited by 1
Corneal Localization for Discerning Faces using Advanced Machine Learning Algorithms of GAN
Shinde, A., Watekar, S., Shekadar, S., Memon, A., & Gharat, S. (2023) · International Conference on Intelligent Data Communication Technologies and Internet of Things (IDCIoT) · doi:10.1109/idciot56793.2023.10052780 · Cited by 1
“[3] In this work, a detection method to recognize these GAN manufactured artificial faces that use the parameter of irregularities in corneal specular highlights to distinguish between the two, synthesized eyes and the real eyes.”
MethodUnmasking the Illusion: Deepfake Detection through MesoNet
Gupta, A., & Pandey, D. (2024) · IEEE International Conference on Computing, Power and Communication Technologies (IC2PCT) · doi:10.1109/ic2pct60090.2024.10486617 · Cited by 1
“Researchers have provided information on using AutoGAN to find deepfakes [9].”
Inversion Image Pairs for Anti-forensics in the Frequency Domain
Pu, H., Yi, X., Yang, B., Zhao, X., & Liu, C. (2024) · Lecture notes in computer science · doi:10.1007/978-981-97-2585-4_13 · Cited by 1
An Efficient Frequency Domain Based Attribution and Detection Network
Zhang, J., Wang, Y., Tohidypour, H. R., & Nasiopoulos, P. (2025) · IEEE Access · doi:10.1109/access.2025.3534829 · Cited by 1
“Zhang et al.’s observation in [24], which showed that GMs’ fingerprints can be clearly observed in frequency domain, was followed by many other approaches that were also based on frequency domain fingerprints for image attribution.”
AI-generated image detectors are surprisingly easy to mislead… for now
Lyu, Z., Xiao, J., Zhang, C., & Lam, K. (2024) · Asia-Pacific Signal and Information Processing Association Annual Summit and Conference (APSIPA ASC) · doi:10.1109/apsipaasc63619.2025.10849104 · Cited by 1
“…variety of research studies have been conducted to generate feature representations that aim to distinguish generated images from real images, including data augmentation [3], frequency domain artifacts [2], neighboring pixel relationship [5], feature encoding [4], and information compression [8].”
Papadimitriou, F. (2025) · SpringerBriefs in applied sciences and technology · doi:10.1007/978-3-031-82136-3_1 · Cited by 1
Image Quality Assessment for Facial Image Authenticity Verification (FIAV)
Bouhamed, S. A., & Msaddak, Y. (2025) · Communications in computer and information science · doi:10.1007/978-981-96-5884-8_15 · Cited by 1
On the Generalisation Capability of Local Surface Frames in Detecting Diffusion-Based Facial Images
Ciamarra, A., Caldelli, R., & Bimbo, A. D. (2025) · IEEE/CVF Winter Conference on Applications of Computer Vision Workshops (WACVW) · doi:10.1109/wacvw65960.2025.00154 · Cited by 2
DeepFake Technology for Breast Cancer Dataset Generation Using Autoencoders and Deep Neural Networks
Anwar, S., Anwer, M. A., & Al-Nadawi, D. (2025) · Lecture notes in computer science · doi:10.1007/978-3-031-88220-3_1 · Cited by 1
Simple Detection of AI-Generated Images based on Noise Correlation
Mallet, A., Méreur, A., Kuribayashi, M., Cogranne, R., & Bas, P. (2025) · International Conference on Advanced Machine Learning and Data Science (AMLDS) · doi:10.1109/amlds63918.2025.11159445 · Cited by 1
“Note that this analysis is commonly used to illustrate the upsampling fingerprint found in both GANs and DMs [6], [7], [11], [33].”
Focal stack based image forgery localization
Huang, Z., Fessler, J. A., & Norris, T. B. (2022) · Applied Optics · doi:10.1364/ao.450654 · Cited by 1
Self-supervised GAN DetectorPreprint
Jeong, Y., Kim, D., Kim, P., Ro, Y., & Choi, J. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2111.06575 · Cited by 1
“Recently, [15] considers utilizing the Discrete Cosine Transform [36] to analyze GAN-related artifacts in frequency space, while [37] provides a detection approach based on the artifacts caused by the up-sampler of GANs.”
MethodFake Image Detection: A Comprehensive ReviewPreprint
Upadhayay, B., & Sahu, S. (2024) · SSRN Electronic Journal · doi:10.2139/ssrn.4779382 · Cited by 1
Synthetic Art Generation and Deepfake Detection: A Study on Jamini Roy Inspired DatasetPreprint
Agrawal, K., & Banerjee, R. (2025) · SSRN Electronic Journal · doi:10.2139/ssrn.5358869 · Cited by 1
Exploring constraints on CycleGAN-based CBCT enhancement for adaptive radiotherapyPreprint
Pai, S. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2110.04659
Pixel-Wise Information in Fake Image Detection
Ngo, N.-K., & Cao, X.-N. (2021) · Communications in computer and information science · doi:10.1007/978-981-16-8062-5_30
Detection of Fake Remote-Sensing Data
Gashnikov, M. V., & Кузнецов, А. В. (2022) · Optical Memory and Neural Networks · doi:10.3103/s1060992x22010052
“However, the authors [14] use wholly forged remote-sensing images to test their detector.”
BackgroundMethodInfluentialAn Overview of Face Image Forgery Detection
He, D., Liu, S., Jin, X., Huang, S., & Jiang, Q. (2022) · Current Chinese Computer Science · doi:10.2174/2665997202666220818122419
Mining Fine-Grain Face Forgery Cues with Fusion Modality
Peng, S., Cai, M., Lu, T., & Liu, X. (2023) · Computers, materials & continua/Computers, materials & continua (Print) · doi:10.32604/cmc.2023.036688
A Methodological Study of Fake Image Creation and Detection Techniques in Multimedia Forensics
Popli, R., Kansal, I., Kumar, R., & Chauhan, R. (2023) · Advances in multimedia and interactive technologies book series · doi:10.4018/978-1-6684-6864-7.ch007
A Scoping Review of GAN-Generated Images Detection
Kit, K. S., Wong, W. K., Chew, I. M., Juwono, F. H., & Sivakumar, S. (2023) · International Conference on Digital Applications, Transformation & Economy (ICDATE) · doi:10.1109/icdate58146.2023.10248679 · Cited by 1
“However, generating ”fake” images using a GAN raises ethical concerns, especially if the images are used to spread false information or manipulate public opinion [3].”
BackgroundBreast Cancer Radiogenomics Data Generation Using Combined Generative Adversarial Networks GANs
Anwar, S., Ali, S. A., Abdulla, D., Omekara, S. D., Mendiola, S., Wright, K., & Muhammed, S. A. (2023) · Communications in computer and information science · doi:10.1007/978-3-031-46335-8_9
Image and video generation via deep learning
Jiang, L. (2023) · doi:10.32657/10356/172067
GAN-generated face detection by color feature and ResNet
Zhang, L. (2023) · International Conference on Algorithms, High Performance Computing and Artificial Intelligence (AHPCAI) · doi:10.1117/12.3011989
“Similarly, Xuan et al. [3] employed data augmentation to enhance generalization, but the progress was limited by the detection algorithm utilized.”
MethodPatterns and Actors of Disinformation: Analysis of Debunked Hoaxes in Spain in 2022
Blanco‐Herrero, D., Castillo-Abdul, B., & Romero-Rodríguez, L. M. (2023) · KOME · doi:10.17646/kome.of.3
“At the same time, misinformation has been studied primarily as a textual problem, but aspects such as deepfakes and images generated by AI ( Yang et al. 2019; Zhang et al. 2019) are also adding new dimensions to the problem.”
BackgroundRobust Training for Deepfake Detection Models Against Disruption-Induced Data Poisoning
Park, J., Ahn, H. E., Park, L. H., & Kwon, T. (2024) · Lecture notes in computer science · doi:10.1007/978-981-99-8024-6_14
Sentinel-2 Image Generation via Stylegan3 Model
Alibani, M., Acito, N., & Corsini, G. U. (2023) · Workshop on Hyperspectral Image and Signal Processing: Evolution in Remote Sensing · doi:10.1109/whispers61460.2023.10430833
“The ability of AI-based generative systems to generate ”fake” images and manipulate content has significant socio-economic implications ([6]), prompting the development of methods suitable for authenticity identification and detection of manipulated data ([7], [8]).”
BackgroundImage Inpainting Forensics Algorithm Based on Dual-Domain Encoder-Decoder Network
Zhang, D., Tan, E., Li, F., Liu, S., Wang, J., & Hu, J. (2024) · Lecture notes in computer science · doi:10.1007/978-981-97-0808-6_6
Pp-Dfd: A Privacy-Preserving Deepfake DetectionPreprint
Jena, R., Das, M., & Singh, P. (2024) · SSRN Electronic Journal · doi:10.2139/ssrn.4779062
Synthetic Media Analysis Using Deep Learning
Vaishnavi Mishra (2024) · INTERANTIONAL JOURNAL OF SCIENTIFIC RESEARCH IN ENGINEERING AND MANAGEMENT · doi:10.55041/ijsrem32494
Improving Facial Recognition: Evaluating CNNs for Detecting Manipulated Images
Bansal, P., Shankar, A., Kumar, P., Vaid, N., & Rajakumar, P. (2024) · International Conference on Reliability, Infocom Technologies and Optimization (Trends and Future Directions) (ICRITO) · doi:10.1109/icrito61523.2024.10522421
“The deepfake image detector developed by the authors of [3] demonstrated encouraging results when trained using the image spectrum.”
MethodCannas, E. D., Mandelli, S., Bestagini, P., & Tubaro, S. (2024) · ACM Workshop on Information Hiding and Multimedia Security (IH&MMSec) · doi:10.1145/3658664.3659644 · Cited by 1
“⇥ ⇥ 46].”
Multi-Frequency-Aware Patch Adversarial Learning for Neural Point Cloud Rendering
Karhade, J., Zhu, H., Chung, K.-S., Tripathy, R. K., Lin, W., & Ang, M. H. (2024) · IEEE International Conference on Advanced Intelligent Mechatronics (AIM) · doi:10.1109/aim55361.2024.10637094
Tran, M., Nguyen, B.-T., Nguyen, V.-L., Nguyen, T., Tran, M.-K., & Do, T.-L. (2024) · International Conference on Multimedia Analysis and Pattern Recognition (MAPR) · doi:10.1109/mapr63514.2024.10661051
“Feature-based methods extract features from the image that indicate forgery and then use these features to classify the image as forged or genuine [6, 9, 19].”
BackgroundUncertainty guided test-time training for face forgery detection
Xu, P., He, Y., Yang, J. Y., & Zhang, S. (2024) · Computer Vision and Image Understanding · doi:10.1016/j.cviu.2024.104185
Denoising of 4D-STEM Dataset using Pix2Pix GAN Algorithm and Artifact Reduction Strategy
Cao, J., Demortière, A., Folastre, N., Oney, G., Das, P. P., & Nicolopoulos, S. (2024) · BIO Web of Conferences · doi:10.1051/bioconf/202412910017
TT-DF: A Large-Scale Diffusion-Based Dataset and Benchmark for Human Body Forgery Detection
Yang, W., Zhang, Z., Zhou, X., Duan, J., & Cao, J. (2024) · Lecture notes in computer science · doi:10.1007/978-981-97-8795-1_29
“Conversely, frequency-aware methods [29,23,48,26,39] represent a paradigm in focusing on the common characteristics of synthetic images, since previous works [47,14] point out that upsampling in generative models can cause synthetic images to deviate from the natural frequency distribution.”
Rethinking the Quality of Synthetic Palm Vein Images from Spectral Analysis
Clarke, C., Salazar-Jurado, E. H., & Hernández-García, R. (2024) · Lecture notes in computer science · doi:10.1007/978-3-031-76604-6_5
AIGC Detection Model Based on Capsule Networks
Zhu, J. (2024) · Applied and Computational Engineering · doi:10.54254/2755-2721/2025.18828
Detecting Face Synthesis Using a Concealed Fusion Model
Leyva, R., Epiphaniou, G., Maple, C. R., & Sánchez, V. (2024) · Communications in computer and information science · doi:10.1007/978-3-031-74627-7_43
Wen, Z., Yin, Z., & Xu, M. (2025) · AI and Optical Data Sciences · doi:10.1117/12.3043727
“GAN artifacts are common and challenging to eliminate[16].”
Impact of Learning Rate on CNN-Based Deepfake Detection
Fatima, A., & Ram, P. K. (2025) · Journal of Artificial Intelligence and Capsule Networks · doi:10.36548/jaicn.2025.1.001
Protecting Images From Manipulations With Deep Optical Signatures
Arias, K., Gomez, P., Hinojosa, C. A., Niebles, J. C., & Argüello, H. (2025) · IEEE Journal of Selected Topics in Signal Processing · doi:10.1109/jstsp.2025.3554136
“For example, authors in [32] propose to extract the median and high frequencies to detect the upsampling artifacts by GANs.”
Liu, Y., Luo, J., Wang, X., Xiao, H., Zhu, K., Nan, Y., & Chen, Y. (2025) · Lecture notes in networks and systems · doi:10.1007/978-981-96-3949-6_8
Detect Fake with Fake: Leveraging Synthetic Data-Driven Representation for Synthetic Image Detection
Otake, H., Fukuhara, Y., Kubotani, Y., & Morishima, S. (2025) · Lecture notes in computer science · doi:10.1007/978-3-031-92648-8_19
Yang, T., & Zhang, Q. (2025) · Signal Image and Video Processing · doi:10.1007/s11760-025-04486-w · Cited by 2
“However, in some individual tasks of CycleGAN, the performance of our method is slightly lower than that of the Freq-spec method [16].”
Attention-Based Refinement and Criss-Cross Mechanisms in GANs for Improved Deepfake Detection
Selvaraj, P., & Belmin, S. A. (2025) · Communications in computer and information science · doi:10.1007/978-3-032-00884-8_26
A Generalization-Enhanced Method for Forged Face Detection Based on High-Frequency Features
Wang, Y., Fan, L., Cui, J., & Zheng, H. (2025) · IEICE Transactions on Communications · doi:10.23919/transcom.2024ebp3212
Face Deepfake Detection Based on Up-Sampling Feature Extraction and CNN
Tang, S. (2025) · Communications in computer and information science · doi:10.1007/978-3-031-94962-3_12
DEVision: A Lightweight Model for Deepfake Video DetectionPreprint
Ghosh, P. S., Paul, R., Koley, S., Kumar, P., Sarkar, D., Chowdhury, R., & Goto, T. (2025) · SSRN Electronic Journal · doi:10.2139/ssrn.5472931
ZKP-StylePatch: Hybrid NFT Anti-counterfeit Framework
Wu, T., Shen, Y. L., Zhang, F., Jiang, Y. H., Lei, Y., Zhou, B., & Fang, J. (2025) · Lecture notes in computer science · doi:10.1007/978-3-032-06176-8_2
Frequency-Domain Acceleration for 3D Generative Adversarial Networks
Jiang, Z., Liu, B., Tang, Y., & Wu, J. (2023) · Journal of Computer-Aided Design & Computer Graphics · doi:10.3724/sp.j.1089.2023.19439
“FAG 通过利用上采样零插入操作后的频域数据重复特 征 (4 部分数据均相同 , 只需要存储 1/4 [14] ) 和共轭 对称的数据特征 ( 只需存储约 1/2 数据 ) 实现减少计 算和访存规模 .”
InfluentialPresent and Future Generalization of Synthetic Image Detectors
Bernabeu-Pérez, P., Lopez-Cuena, E., & García-Gasulla, D. (2025) · Lecture notes in computer science · doi:10.1007/978-3-032-06109-6_1 · Cited by 4
Model Synthesis for Zero-Shot Model Attribution
Yang, T., Wang, D., Cao, J., & Xu, C. (2025) · IEEE Transactions on Multimedia · doi:10.1109/tmm.2025.3607778 · Cited by 3
“Zhang et al. [16] showed that the upsampling layer causes a periodic grid pattern on the spectrum.”
InfluentialEvading Deepfake Detectors via Adversarially Degrading and Restoring Forged Images
Shi, Z., Lin, C., Zhao, Z., Peer, P., & Shen, C. (2025) · IEEE International Conference on Multimedia and Expo (ICME) · doi:10.1109/icme59968.2025.11210143
“At the spectral level, generative models introduce frequency artifacts, such as high-frequency components [1].”
DBE: Dual Branch re-Extraction for Unseen Diffusion-Generated Image Detection
Cai, S., Liu, L., Kuai, Z., Kuang, L., & Zhang, L. (2025) · IEEE International Conference on Multimedia and Expo (ICME) · doi:10.1109/icme59968.2025.11209962
“Zhang et al. [12] and Frank et al. [13] researched artifacts in the frequency spectrum of GAN-generated images.”
AIGC Video Detection Based on Missing Motion Details
Cao, Y., Tian, Y., Huang, L., Bao, Y., & Li, Q. (2025) · Lecture notes in computer science · doi:10.1007/978-981-95-3729-7_5
FRADet: Frequency Refinement Augmentation Diffusion Images Detector
Li, D., & Zhang, H. (2025) · International Conference on Internet of Things, Artificial Intelligence and Mechanical Automation (IoTAIMA) · doi:10.1109/iotaima66468.2025.11212709
“Some studies [7,11] also found that there are obvious visual artifacts in images generated by GAN, prompting detection methods to the perspective of frequency domain.”
CLIP-Enhanced Linear Alignment for AI-Generated Image Detection Using Latent Cross-Attention
Wang, Y., & Sun, T. (2025) · International Joint Conference on Neural Networks (IJCNN) · doi:10.1109/ijcnn64981.2025.11228586
“Some works [18], [19] have found that converting the spatial information of GAN-generated images into frequency-domain information allows for detecting anomalies in the frequency domain to identify fake images.”
Wen, L., He, X., Fu, Z., Yang, T., Wang, L., & Wang, Q. (2025) · The Journal of Supercomputing · doi:10.1007/s11227-025-08043-7
“Zhang et al [2] pointed out the checkerboard artifacts introduced by the transposed convolution operation, but the existing methods lack a targeted feature enhancement mechanism, which makes it difficult to stably extract such anomalies in complex scenes.”
InfluentialSelective Recompression for Image Forgery Localisation in non-JPEG Images
Abraham, W., Shekar, B. H., & Pilar, B. (2025) · IEEE International Conference on Distributed Computing, VLSI, Electrical Circuits and Robotics (DISCOVER) · doi:10.1109/discover66922.2025.11258986
“Zhang et al. (2019) [8] applied GANs for anomaly detection, training the network to learn the distribution of authentic images and identify anomalies”
JPEG-compression agnostic identification of generative art using explainable spatial domain features
Kohli, V., Shukla, J., Sharma, H. K., & Khatri, N. (2025) · Array · doi:10.1016/j.array.2025.100635
Frequency domain discrimination and double loss function optimization of image generation for GANs
Ye, X., Sui, M., Zeng, M., Han, Z., & Wang, H. (2025) · The Journal of Supercomputing · doi:10.1007/s11227-025-08105-w · Cited by 1
“Although some scholars [6–11] have shifted their focus to the frequency domain, they have not deeply investigated the generator’s decoding methods (i.e., convolution types).”
FFRNet: Exploring Forgery Cues Via Adaptive Feature Scaling for Deepfake Detection
Jin, P., & Xu, S. (2025) · International Conference on Computer Vision and Data Mining (ICCVDM) · doi:10.1109/iccvdm66874.2025.11290338
“AutoGAN [33] took a novel approach by building a spectral-domain discriminator, effectively overcoming limitations of purely spatial-domain methods.”
Detection of AI-Generated Facial Images Using Convolutional Neural Networks
Rayadi, A. M., Suparwito, H., & Nugaliyadde, A. (2026) · E3S Web of Conferences · doi:10.1051/e3sconf/202668702010
In-depth survey of the technologies to combat disinformation
Rodríguez-González, L. I., Arias-González, M., Armenteros-Cosme, P., Mateos-Plaza, S., & Hernández, J. C. (2025) · IEEE International Conference on Frontiers Technology of Information and Computer (ICFTIC) · doi:10.1109/icftic68075.2025.11324595
“In [48], a GAN simulator is introduced that generates synthetic content to train a classifier, improving its ability to identify deepfakes.”
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
