mlsanityData to 14 Oct 2026

High Frequency Component Helps Explain the Generalization of Convolutional Neural Networks

arXiv 1905.13545 ↗First posted 28 May 2019cs.CV cs.LGcited by 310

Not on the map. Why?

Abstract (arXiv)

We investigate the relationship between the frequency spectrum of image data and the generalization behavior of convolutional neural networks (CNN). We first notice CNN's ability in capturing the high-frequency components of images. These high-frequency components are almost imperceptible to a human. Thus the observation leads to multiple hypotheses that are related to the generalization behaviors of CNN, including a potential explanation for adversarial examples, a discussion of CNN's trade-off between robustness and accuracy, and some evidence in understanding training heuristics.

Key facts. Cited by 310 papers in the map (data through 14 Oct 2026). First posted on arXiv 28 May 2019.

Builds on

No reference list yet.

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2021: 50; 2022: 62; 2023: 58; 2024: 55; 2025: 45; 2026: 40.

310 papers.

Sort:

By area:

  1. 2021Generalized Out-of-Distribution Detection: A Surveycited by 802
  2. 2021ILVR: Conditioning Method for Denoising Diffusion Probabilistic Modelscited by 635
  3. 2021Spatial-Phase Shallow Learning: Rethinking Face Forgery Detection in Frequency Domaincited by 244
  4. 2023FreeU: Free Lunch in Diffusion U-Netcited by 196
  5. 2021Unlearnable Examples: Making Personal Data Unexploitablecited by 193
  6. 2021FSDR: Frequency Space Domain Randomization for Domain Generalizationcited by 164
  7. 2021Reading Race: AI Recognises Patient's Racial Identity In Medical Imagescited by 156
  8. 2022Frequency Domain Model Augmentation for Adversarial Attackcited by 130
  9. 2021Measure and Improve Robustness in NLP Models: A Surveycited by 107
  10. 2021On the Adversarial Robustness of Vision Transformerscited by 106
  11. 2022Inception Transformercited by 99
  12. 2021Towards Interpreting and Mitigating Shortcut Learning Behavior of NLU Modelscited by 78
38 more
  1. 2021When Does Contrastive Learning Preserve Adversarial Robustness from Pretraining to Finetuning?cited by 77
  2. 2022Frequency-driven Imperceptible Adversarial Attack on Semantic Similaritycited by 77
  3. 2022Improving Vision Transformers by Revisiting High-frequency Componentscited by 76
  4. 2023A Pilot Study of Query-Free Adversarial Attack against Stable Diffusioncited by 76
  5. 2024Wavelet Convolutions for Large Receptive Fieldscited by 74
  6. 2021Amplitude-Phase Recombination: Rethinking Robustness of Convolutional Neural Networks in Frequency Domaincited by 71
  7. 2021On the Frequency Bias of Generative Modelscited by 60
  8. 2025DeCo: Frequency-Decoupled Pixel Diffusion for End-to-End Image Generationcited by 60
  9. 2025Frequency-Aligned Knowledge Distillation for Lightweight Spatiotemporal Forecastingcited by 55
  10. 2024Frequency-aware Feature Fusion for Dense Image Predictioncited by 48
  11. 2023Rethinking Out-of-distribution (OOD) Detection: Masked Image Modeling is All You Needcited by 46
  12. 2023Adaptive Frequency Filters As Efficient Global Token Mixerscited by 45
  13. 2021Robust Contrastive Learning Using Negative Samples with Diminished Semanticscited by 43
  14. 2022Deep Frequency Filtering for Domain Generalizationcited by 43
  15. 2022EfficientTrain: Exploring Generalized Curriculum Learning for Training Visual Backbonescited by 43
  16. 2024Efficient Real-world Image Super-Resolution Via Adaptive Directional Gradient Convolutioncited by 37
  17. 2021A Little Robustness Goes a Long Way: Leveraging Robust Features for Targeted Transfer Attackscited by 34
  18. 2024Frequency-Adaptive Dilated Convolution for Semantic Segmentationcited by 34
  19. 2022The Devil is in the Frequency: Geminated Gestalt Autoencoder for Self-Supervised Visual Pre-Trainingcited by 33
  20. 2023Downstream-agnostic Adversarial Examplescited by 32
  21. 2021Reliably fast adversarial training via latent adversarial perturbationcited by 29
  22. 2021WaveCNet: Wavelet Integrated CNNs to Suppress Aliasing Effect for Noise-Robust Image Classificationcited by 29
  23. 2021A Frequency Perspective of Adversarial Robustnesscited by 28
  24. 2022Privacy-Preserving Face Recognition with Learnable Privacy Budgets in Frequency Domaincited by 28
  25. 2023What do neural networks learn in image classification? A frequency shortcut perspectivecited by 28
  26. 2023Spatial-frequency channels, shape bias, and adversarial robustnesscited by 28
  27. 2021Enhancing Adversarial Robustness via Test-time Transformation Ensemblingcited by 27
  28. 2022Gradient Methods Provably Converge to Non-Robust Networkscited by 27
  29. 2025More Thought, Less Accuracy? On the Dual Nature of Reasoning in Vision-Language Modelscited by 27
  30. 2021Learning domain-agnostic visual representation for computational pathology using medically-irrelevant style transfer augmentationcited by 26
  31. 2021Towards End-to-End Image Compression and Analysis with Transformerscited by 26
  32. 2023Decompose, Adjust, Compose: Effective Normalization by Playing with Frequency for Domain Generalizationcited by 26
  33. 2021Taxonomy of Machine Learning Safety: A Survey and Primercited by 25
  34. 2024LEAD: Learning Decomposition for Source-free Universal Domain Adaptationcited by 25
  35. 2021Removing Undesirable Feature Contributions Using Out-of-Distribution Datacited by 23
  36. 2022Out-of-distribution Detection via Frequency-regularized Generative Modelscited by 23
  37. 2023The Double-Edged Sword of Implicit Bias: Generalization vs. Robustness in ReLU Networkscited by 23
  38. 2024Can We Talk Models Into Seeing the World Differently?cited by 23

50 most cited of 310 All 310 (JSON)