mlsanityData to 7 Oct 2026

Exploring The Effect of High-frequency Components in GANs Training

arXiv 2103.11093 ↗First posted 20 Mar 2021cs.CVcited by 10

Not on the map. Why?

Abstract (arXiv)

Generative Adversarial Networks (GANs) have the ability to generate images that are visually indistinguishable from real images. However, recent studies have revealed that generated and real images share significant differences in the frequency domain. In this paper, we explore the effect of high-frequency components in GANs training. According to our observation, during the training of most GANs, severe high-frequency differences make the discriminator focus on high-frequency components excessively, which hinders the generator from fitting the low-frequency components that are important for learning images' content. Then, we propose two simple yet effective frequency operations for eliminating the side effects caused by high-frequency differences in GANs training: High-Frequency Confusion (HFC) and High-Frequency Filter (HFF). The proposed operations are general and can be applied to most existing GANs with a fraction of the cost. The advanced performance of the proposed operations is verified on multiple loss functions, network architectures, and datasets. Specifically, the proposed HFF achieves significant improvements of 42.5% FID on CelebA (128*128) unconditional generation based on SNGAN, 30.2% FID on CelebA unconditional generation based on SSGAN, and 69.3% FID on CelebA unconditional generation based on InfoMAXGAN.

Key facts. Cited by 10 papers in the map (data through 7 Oct 2026). Builds on 49 of its 74 references that are in the map. First posted on arXiv 20 Mar 2021.

Builds on

49 of 74 references.

References · 49

  1. 2020A Simple Framework for Contrastive Learning of Visual Representationscited by 11,979
  2. 2018Representation Learning with Contrastive Predictive Codingcited by 7,662
  3. 2019Momentum Contrast for Unsupervised Visual Representation Learningcited by 7,482
  4. 2014Explaining and Harnessing Adversarial Examplescited by 7,287
  5. 2018A Style-Based Generator Architecture for Generative Adversarial Networkscited by 5,986
  6. 2019Analyzing and Improving the Image Quality of StyleGANcited by 3,580
  7. 2017Progressive Growing of GANs for Improved Quality, Stability, and Variationcited by 2,967
  8. 2016Improved Techniques for Training GANscited by 2,938
  9. 2017Improved Training of Wasserstein GANscited by 2,223
  10. 2020Fourier Features Let Networks Learn High Frequency Functions in Low Dimensional Domainscited by 2,170
  11. 2018Spectral Normalization for Generative Adversarial Networkscited by 1,618
  12. 2015LSUN: Construction of a Large-scale Image Dataset using Deep Learning with Humans in the Loopcited by 1,350
37 more
  1. 2018On the Spectral Bias of Neural Networkscited by 1,321
  2. 2018Demystifying MMD GANscited by 1,268
  3. 2020Training Generative Adversarial Networks with Limited Datacited by 1,184
  4. 2016Least Squares Generative Adversarial Networkscited by 1,000
  5. 2019Interpreting the Latent Space of GANs for Semantic Face Editingcited by 691
  6. 2017Wasserstein GANcited by 635
  7. 2016f-GAN: Training Generative Neural Samplers using Variational Divergence Minimizationcited by 504
  8. 2020Leveraging Frequency Analysis for Deep Fake Image Recognitioncited by 501
  9. 2020Differentiable Augmentation for Data-Efficient GAN Trainingcited by 361
  10. 2019Frequency Principle: Fourier Analysis Sheds Light on Deep Neural Networkscited by 360
  11. 2019High Frequency Component Helps Explain the Generalization of Convolutional Neural Networkscited by 310
  12. 2019Detecting and Simulating Artifacts in GAN Fake Imagescited by 293
  13. 2020Watch your Up-Convolution: CNN Based Generative Deep Neural Networks are Failing to Reproduce Spectral Distributionscited by 278
  14. 2017Geometric GANcited by 232
  15. 2019Image Processing Using Multi-Code GAN Priorcited by 185
  16. 2018Training behavior of deep neural network in frequency domaincited by 174
  17. 2019Towards Understanding the Spectral Bias of Deep Learningcited by 173
  18. 2019Fourier Spectrum Discrepancies in Deep Network Generated Imagescited by 138
  19. 2019The Convergence Rate of Neural Networks for Learned Functions of Different Frequenciescited by 134
  20. 2018Self-Supervised GANs via Auxiliary Rotation Losscited by 94
  21. 2020Image Augmentations for GAN Trainingcited by 78
  22. 2021Training GANs with Stronger Augmentations via Contrastive Discriminatorcited by 45
  23. 2018Understanding training and generalization in deep learning by Fourier analysiscited by 43
  24. 2017Perceptual Adversarial Networks for Image-to-Image Transformationcited by 42
  25. 2020SSD-GAN: Measuring the Realness in the Spatial and Spectral Domainscited by 38
  26. 2020InfoMax-GAN: Improved Adversarial Image Generation via Information Maximization and Contrastive Learningcited by 37
  27. 2020Spatial Frequency Bias in Convolutional Generative Adversarial Networkscited by 35
  28. 2019Conditional LSTM-GAN for Melody Generation from Lyricscited by 31
  29. 2017CM-GANs: Cross-modal Generative Adversarial Networks for Common Representation Learningcited by 27
  30. 2020A Systematic Survey of Regularization and Normalization in GANscited by 27
  31. 2019PEPSI++: Fast and Lightweight Network for Image Inpaintingcited by 24
  32. 2019Explicitizing an Implicit Bias of the Frequency Principle in Two-layer Neural Networkscited by 20
  33. 2020A New Perspective on Stabilizing GANs training: Direct Adversarial Trainingcited by 12
  34. 2019Synthesizing facial photometries and corresponding geometries using generative adversarial networkscited by 11
  35. 2020LT-GAN: Self-Supervised GAN with Latent Transformation Detectioncited by 11
  36. 2020Learning to fool the speaker recognitioncited by 5
  37. 2018Frequency Principle in Deep Learning with General Loss Functions and Its Potential Applicationcited by 3

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2023: 2; 2024: 1; 2025: 6; 2026: 1.

10 papers.

  1. 2025DeCo: Frequency-Decoupled Pixel Diffusion for End-to-End Image Generationcited by 55
  2. 2025Is Artificial Intelligence Generated Image Detection a Solved Problem?cited by 49
  3. 2023Efficient Backdoor Attacks for Deep Neural Networks in Real-world Scenarioscited by 7
  4. 2023Peer is Your Pillar: A Data-unbalanced Conditional GANs for Few-shot Image Generationcited by 5
  5. 2024One-shot Generative Domain Adaptation in 3D GANscited by 3
  6. 2025DisCoPatch: Taming Adversarially-driven Batch Statistics for Improved Out-of-Distribution Detectioncited by 1
  7. 2025Zero-Shot Visual Deepfake Detection: Can AI Predict and Prevent Fake Content Before It's Created?cited by 1
  8. 2025AdverX-Ray: Ensuring X-Ray Integrity Through Frequency-Sensitive Adversarial VAEscited by 0
  9. 2025Structure-Preserving Unpaired Image Translation to Photometrically Calibrate JunoCam with Hubble Datacited by 0
  10. 2026Close Shortcut Wins Long: Seeking Diverse and Stable Generators for Data-Free Knowledge Distillationcited by 0