mlsanityData to 7 Oct 2026

A Study and Comparison of Human and Deep Learning Recognition Performance Under Visual Distortions

arXiv 1705.02498 ↗First posted 6 May 2017cs.CVcited by 107

Not on the map. Why?

Abstract (arXiv)

Deep neural networks (DNNs) achieve excellent performance on standard classification tasks. However, under image quality distortions such as blur and noise, classification accuracy becomes poor. In this work, we compare the performance of DNNs with human subjects on distorted images. We show that, although DNNs perform better than or on par with humans on good quality images, DNN performance is still much lower than human performance on distorted images. We additionally find that there is little correlation in errors between DNNs and human subjects. This could be an indication that the internal representation of images are different between DNNs and the human visual system. These comparisons with human performance could be used to guide future development of more robust DNNs.

Key facts. Cited by 107 papers in the map (data through 7 Oct 2026). First posted on arXiv 6 May 2017.

Builds on

No reference list yet.

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2021: 32; 2022: 23; 2023: 26; 2024: 12; 2025: 9; 2026: 5.

107 papers.

Sort:

By area:

  1. 2021Learning Transferable Visual Models From Natural Language Supervisioncited by 33,456
  2. 2021Partial success in closing the gap between human and machine visioncited by 185
  3. 20223D Common Corruptions and Data Augmentationcited by 120
  4. 2021On Interaction Between Augmentations and Corruptions in Natural Corruption Robustnesscited by 98
  5. 2022Vision Models Are More Robust And Fair When Pretrained On Uncurated Images Without Supervisioncited by 80
  6. 2022Human alignment of neural network representationscited by 74
  7. 2021On the effectiveness of adversarial training against common corruptionscited by 71
  8. 2023On the Design Fundamentals of Diffusion Models: A Surveycited by 59
  9. 2024A Comprehensive Survey of Foundation Models in Medicinecited by 56
  10. 2021If your data distribution shifts, use self-learningcited by 36
  11. 2022Enhancing Diffusion-Based Image Synthesis with Robust Classifier Guidancecited by 33
  12. 2021A Review of Adversarial Attack and Defense for Classification Methodscited by 29
38 more
  1. 2023MMG-Ego4D: Multi-Modal Generalization in Egocentric Action Recognitioncited by 26
  2. 2021Robustness via Cross-Domain Ensemblescited by 23
  3. 2022Do Perceptually Aligned Gradients Imply Adversarial Robustness?cited by 15
  4. 2023Self-attention in Vision Transformers Performs Perceptual Grouping, Not Attentioncited by 15
  5. 2023Neural Architecture Design and Robustness: A Datasetcited by 15
  6. 2024LLM Whisperer: An Inconspicuous Attack to Bias LLM Responsescited by 15
  7. 2022Progress and limitations of deep networks to recognize objects in unusual posescited by 13
  8. 2021Comparing object recognition in humans and deep convolutional neural networks -- An eye tracking studycited by 12
  9. 2021Anti-aliasing Deep Image Classifiers using Novel Depth Adaptive Blurring and Activation Functioncited by 11
  10. 2021DeepAdversaries: Examining the Robustness of Deep Learning Models for Galaxy Morphology Classificationcited by 10
  11. 2023Rapid Network Adaptation: Learning to Adapt Neural Networks Using Test-Time Feedbackcited by 10
  12. 2021Rethinking Natural Adversarial Examples for Classification Modelscited by 9
  13. 2021Benchmarks for Corruption Invariant Person Re-identificationcited by 9
  14. 2024MomentumSMoE: Integrating Momentum into Sparse Mixture of Expertscited by 9
  15. 2021Closer Look at the Uncertainty Estimation in Semantic Segmentation under Distributional Shiftcited by 8
  16. 2021Failure Modes of Domain Generalization Algorithmscited by 8
  17. 2022Test-time Batch Normalizationcited by 8
  18. 2023DeepAstroUDA: Semi-Supervised Universal Domain Adaptation for Cross-Survey Galaxy Morphology Classification and Anomaly Detectioncited by 8
  19. 2024Imperceptible Protection against Style Imitation from Diffusion Modelscited by 8
  20. 2021Inverse Problems Leveraging Pre-trained Contrastive Representationscited by 7
  21. 2022A Survey on Infrared Image and Video Setscited by 7
  22. 2022A Survey on AI Sustainability: Emerging Trends on Learning Algorithms and Research Challengescited by 7
  23. 2022Adversarial Framework with Certified Robustness for Time-Series Domain via Statistical Featurescited by 7
  24. 2022Runtime Monitoring for Out-of-Distribution Detection in Object Detection Neural Networkscited by 7
  25. 2021Contemplating real-world object classificationcited by 6
  26. 2022Pyramid Frequency Network with Spatial Attention Residual Refinement Module for Monocular Depth Estimationcited by 6
  27. 2023Generalizable Lightweight Proxy for Robust NAS against Diverse Perturbationscited by 6
  28. 2023Instilling Inductive Biases with Subnetworkscited by 6
  29. 2021CheXternal: Generalization of Deep Learning Models for Chest X-ray Interpretation to Photos of Chest X-rays and External Clinical Settingscited by 5
  30. 2021Combining Different V1 Brain Model Variants to Improve Robustness to Image Corruptions in CNNscited by 5
  31. 2022Does lossy image compression affect racial bias within face recognition?cited by 5
  32. 2022Data Models for Dataset Drift Controls in Machine Learning With Optical Imagescited by 5
  33. 2023Class-Conditioned Transformation for Enhanced Robust Image Classificationcited by 5
  34. 2023Training on Foveated Images Improves Robustness to Adversarial Attackscited by 5
  35. 2021Towards Corruption-Agnostic Robust Domain Adaptationcited by 4
  36. 2023Investigating the Corruption Robustness of Image Classifiers with Random Lp-norm Corruptionscited by 4
  37. 2023Cycle Consistency-based Uncertainty Quantification of Neural Networks in Inverse Imaging Problemscited by 4
  38. 2024Algorithmic Ways of Seeing: Using Object Detection to Facilitate Art Explorationcited by 4

50 most cited of 107 All 107 (JSON)