Data to 14 Oct 2026

CycleGAN, a Master of Steganography

arXiv 1712.02950 ↗First posted 8 Dec 2017cs.CV cs.LG stat.MLcited by 65

Not on the map. Why?

Abstract (arXiv)

CycleGAN (Zhu et al. 2017) is one recent successful approach to learn a transformation between two image distributions. In a series of experiments, we demonstrate an intriguing property of the model: CycleGAN learns to "hide" information about a source image into the images it generates in a nearly imperceptible, high-frequency signal. This trick ensures that the generator can recover the original sample and thus satisfy the cyclic consistency requirement, while the generated image remains realistic. We connect this phenomenon with adversarial attacks by viewing CycleGAN's training procedure as training a generator of adversarial examples and demonstrate that the cyclic consistency loss causes CycleGAN to be especially vulnerable to adversarial attacks.

Key facts. Cited by 65 papers in the map (data through 14 Oct 2026). First posted on arXiv 8 Dec 2017.

Builds on

No reference list yet.

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2019: 1; 2020: 0; 2021: 15; 2022: 15; 2023: 12; 2024: 5; 2025: 10; 2026: 7.

65 papers.

Sort:

By area:

2022The Alignment Problem from a Deep Learning Perspectivecited by 258
2019Variational Mixture-of-Experts Autoencoders for Multi-Modal Deep Generative Modelscited by 178
2022Learning to generate line drawings that convey geometry and semanticscited by 83
2023Preventing Language Models From Hiding Their Reasoningcited by 65
2023LEO: Generative Latent Image Animator for Human Video Synthesiscited by 40
2021High-Fidelity and Arbitrary Face Editingcited by 35
2021DeFlow: Learning Complex Image Degradations from Unpaired Data with Conditional Flowscited by 30
2021Controllable Generation from Pre-trained Language Models via Inverse Promptingcited by 29
2021Segmentation-Renormalized Deep Feature Modulation for Unpaired Image Harmonizationcited by 24
2022FUN-SIS: a Fully UNsupervised approach for Surgical Instrument Segmentationcited by 20
2022Learn from Unpaired Data for Image Restoration: A Variational Bayes Approachcited by 18
2022Synthetic ID Card Image Generation for Improving Presentation Attack Detectioncited by 18

38 more

2024StegoGAN: Leveraging Steganography for Non-Bijective Image-to-Image Translationcited by 18
2022Generative Steganography Networkcited by 15
2023Dataset Condensation via Generative Modelcited by 15
2025Detection of AI Deepfake and Fraud in Online Payments Using GAN-Based Modelscited by 14
2023Inverse Rendering of Translucent Objects using Physical and Neural Rendererscited by 9
2023Exploring Semantic Consistency in Unpaired Image Translation to Generate Data for Surgical Applicationscited by 9
2021Towards Better Adversarial Synthesis of Human Images from Textcited by 7
2025CYCLE-INSTRUCT: Fully Seed-Free Instruction Tuning via Dual Self-Training and Cycle Consistencycited by 6
2025A Chain-of-thought Reasoning Breast Ultrasound Dataset Covering All Histopathology Categoriescited by 6
2021Generating Synthetic Handwritten Historical Documents With OCR Constrained GANscited by 5
2022Quality Metric Guided Portrait Line Drawing Generation from Unpaired Training Datacited by 5
2021Self adversarial attack as an augmentation method for immunohistochemical stainingscited by 4
2021Unsupervised Detection of Cancerous Regions in Histology Imagery using Image-to-Image Translationcited by 4
2022CLTS-GAN: Color-Lighting-Texture-Specular Reflection Augmentation for Colonoscopycited by 4
2024A Pattern Language for Machine Learning Taskscited by 4
2025VaseVQA-3D: Benchmarking 3D VLMs on Ancient Greek Potterycited by 4
2021Style Agnostic 3D Reconstruction via Adversarial Style Transfercited by 3
2022Policy Gradient Stock GAN for Realistic Discrete Order Data Generation in Financial Marketscited by 3
2022Generative Adversarial Networks and Other Generative Modelscited by 3
2025Out of Control -- Why Alignment Needs Formal Control Theory (and an Alignment Control Stack)cited by 3
2026AI Finds A Waycited by 3
2021Generative Network-Based Reduced-Order Model for Prediction, Data Assimilation and Uncertainty Quantificationcited by 2
2021AD-GAN: End-to-end Unsupervised Nuclei Segmentation with Aligned Disentangling Trainingcited by 2
2022Improving Performance of Semantic Segmentation CycleGANs by Noise Injection into the Latent Segmentation Spacecited by 2
2023Standardized CycleGAN training for unsupervised stain adaptation in invasive carcinoma classification for breast histopathologycited by 2
2023Semi-supervised Cycle-GAN for face photo-sketch translation in the wildcited by 2
2023A Novel Deep Learning Technique for Morphology Preserved Fetal ECG Extraction from Mother ECG using 1D-CycleGANcited by 2
2024Texture-AD: An Anomaly Detection Dataset and Benchmark for Real Algorithm Developmentcited by 2
2024Risk Sources and Risk Management Measures in Support of Standards for General-Purpose AI Systemscited by 2
2026Train the Model, Not the Reader: Decodability Supervision for Verifiable Activation Explanationscited by 2
2021Disentangled Unsupervised Image Translation via Restricted Information Flowcited by 1
2022NeuRegenerate: A Framework for Visualizing Neurodegenerationcited by 1
2023Focus on Content not Noise: Improving Image Generation for Nuclei Segmentation by Suppressing Steganography in CycleGANcited by 1
2023Deep Image Semantic Communication Model for Artificial Intelligent Internet of Thingscited by 1
2025Single-image Reflectance and Transmittance Estimation from Any Flatbed Scannercited by 1
20252D Instance Editing in 3D Spacecited by 1
2026HydraPrompt: An Adaptive and Asymmetric Framework of Vision-Language Models for Synthetic Image Detectioncited by 1
2021Self-Supervised Equivariant Scene Synthesis from Videocited by 0

50 most cited of 65 All 65 (JSON)