mlsanityData to 7 Oct 2026

Leveling Down in Computer Vision: Pareto Inefficiencies in Fair Deep Classifiers

arXiv 2203.04913 ↗First posted 9 Mar 2022cs.CV cs.LGcited by 53

Not on the map. Why?

Abstract (arXiv)

Algorithmic fairness is frequently motivated in terms of a trade-off in which overall performance is decreased so as to improve performance on disadvantaged groups where the algorithm would otherwise be less accurate. Contrary to this, we find that applying existing fairness approaches to computer vision improve fairness by degrading the performance of classifiers across all groups (with increased degradation on the best performing groups). Extending the bias-variance decomposition for classification to fairness, we theoretically explain why the majority of fairness classifiers designed for low capacity models should not be used in settings involving high-capacity models, a scenario common to computer vision. We corroborate this analysis with extensive experimental support that shows that many of the fairness heuristics used in computer vision also degrade performance on the most disadvantaged groups. Building on these insights, we propose an adaptive augmentation strategy that, uniquely, of all methods tested, improves performance for the disadvantaged groups.

Key facts. Cited by 53 papers in the map (data through 7 Oct 2026). Builds on 46 of its 87 references that are in the map. First posted on arXiv 9 Mar 2022.

Builds on

46 of 87 references.

References · 46

  1. 2015Deep Residual Learning for Image Recognitioncited by 51,809
  2. 2017mixup: Beyond Empirical Risk Minimizationcited by 5,155
  3. 2014Deep Learning Face Attributes in the Wildcited by 3,679
  4. 2016Equality of Opportunity in Supervised Learningcited by 2,000
  5. 2014Unsupervised Domain Adaptation by Backpropagationcited by 1,906
  6. 2016Understanding deep learning requires rethinking generalizationcited by 1,657
  7. 2016Playing for Data: Ground Truth from Computer Gamescited by 816
  8. 2019Interpreting the Latent Space of GANs for Semantic Face Editingcited by 691
  9. 2018A Reductions Approach to Fair Classificationcited by 585
  10. 2021Pervasive Label Errors in Test Sets Destabilize Machine Learning Benchmarkscited by 388
  11. 2018Learning Adversarially Fair and Transferable Representationscited by 353
  12. 2016Adversarial Feature Learningcited by 350
34 more
  1. 2017Fairness in Criminal Justice Risk Assessments: The State of the Artcited by 315
  2. 2018Learning Not to Learn: Training Deep Neural Networks with Biased Datacited by 255
  3. 2019Towards Fairness in Visual Recognition: Effective Strategies for Bias Mitigationcited by 252
  4. 2018Women also Snowboard: Overcoming Bias in Captioning Modelscited by 236
  5. 2018Empirical Risk Minimization under Fairness Constraintscited by 226
  6. 2016Adversarially Learned Inferencecited by 216
  7. 2015Censoring Representations with an Adversarycited by 191
  8. 2017Data Decisions and Theoretical Implications when Adversarially Learning Fair Representationscited by 165
  9. 2019REPAIR: Removing Representation Bias by Dataset Resamplingcited by 157
  10. 2020Adversarial Latent Autoencoderscited by 155
  11. 2020Minimax Pareto Fairness: A Multi Objective Perspectivecited by 152
  12. 2019Fairness in Recommendation Ranking through Pairwise Comparisonscited by 127
  13. 2020Fair Attribute Classification through Latent Space De-biasingcited by 125
  14. 2018Turning a Blind Eye: Explicit Removal of Biases and Variation from Deep Neural Network Embeddingscited by 114
  15. 2019Fair Generative Modeling via Weak Supervisioncited by 114
  16. 2021Fair Mixup: Fairness via Interpolationcited by 109
  17. 2020Rethinking Bias-Variance Trade-off for Generalization of Neural Networkscited by 107
  18. 2019Diversity in Facescited by 77
  19. 2019Bias Correction of Learned Generative Models using Likelihood-Free Importance Weightingcited by 70
  20. 2016Satisfying Real-world Goals with Dataset Constraintscited by 68
  21. 2020Towards causal benchmarking of bias in face analysis algorithmscited by 64
  22. 2019Discrimination in the Age of Algorithmscited by 51
  23. 2020Analysis of Gender Inequality In Face Recognition Accuracycited by 47
  24. 2020Fair Bayesian Optimizationcited by 45
  25. 2018FNNC: Achieving Fairness through Neural Networkscited by 43
  26. 2020Active Sampling for Min-Max Fairnesscited by 33
  27. 2021Evaluating Fairness of Machine Learning Models Under Uncertain and Incomplete Informationcited by 33
  28. 2018Fairness GANcited by 32
  29. 2017Empirically Analyzing the Effect of Dataset Biases on Deep Face Recognition Systemscited by 28
  30. 2020How fair can we go in machine learning? Assessing the boundaries of fairness in decision treescited by 22
  31. 2019Tackling Algorithmic Bias in Neural-Network Classifiers using Wasserstein-2 Regularizationcited by 18
  32. 2021InvGAN: Invertible GANscited by 7
  33. 2021Pairwise Fairness for Ordinal Regressioncited by 6
  34. 2021Matched sample selection with GANs for mitigating attribute confoundingcited by 4

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2022: 5; 2023: 21; 2024: 13; 2025: 10; 2026: 4.

53 papers.

Sort:

By area:

  1. 2024FairCLIP: Harnessing Fairness in Vision-Language Learningcited by 69
  2. 2022MEDFAIR: Benchmarking Fairness for Medical Imagingcited by 66
  3. 2022Assaying Out-Of-Distribution Generalization in Transfer Learningcited by 53
  4. 2023Harvard Glaucoma Fairness: A Retinal Nerve Disease Dataset for Fairness Learning and Fair Identity Normalizationcited by 31
  5. 2023Benchmarking Algorithmic Bias in Face Recognition: An Experimental Approach Using Synthetic Faces and Human Evaluationcited by 25
  6. 2023FairTune: Optimizing Parameter Efficient Fine Tuning for Fairness in Medical Image Analysiscited by 23
  7. 2023FairAdaBN: Mitigating unfairness with adaptive batch normalization and its application to dermatological disease classificationcited by 18
  8. 2023Evaluating the Fairness of Deep Learning Uncertainty Estimates in Medical Image Analysiscited by 15
  9. 2023Image retrieval outperforms diffusion models on data augmentationcited by 14
  10. 2024Slicing Through Bias: Explaining Performance Gaps in Medical Image Analysis using Slice Discovery Methodscited by 14
  11. 2023Are Sex-based Physiological Differences the Cause of Gender Bias for Chest X-ray Diagnosis?cited by 13
  12. 2023FairVision: Equitable Deep Learning for Eye Disease Screening via Fair Identity Scalingcited by 13
38 more
  1. 2023On (assessing) the fairness of risk score modelscited by 12
  2. 2024OxonFair: A Flexible Toolkit for Algorithmic Fairnesscited by 12
  3. 2024Fairness and Bias Mitigation in Computer Vision: A Surveycited by 12
  4. 2024Robust image representations with counterfactual contrastive learningcited by 12
  5. 2023The Role of Subgroup Separability in Group-Fair Medical Image Classificationcited by 11
  6. 2023No Fair Lunch: A Causal Perspective on Dataset Bias in Machine Learning for Medical Imagingcited by 11
  7. 2024Forte : Finding Outliers with Representation Typicality Estimationcited by 9
  8. 2022FairGrad: Fairness Aware Gradient Descentcited by 8
  9. 2024Rethinking Fair Representation Learning for Performance-Sensitive Taskscited by 8
  10. 2023Are demographically invariant models and representations in medical imaging fair?cited by 6
  11. 2025Robust ML Auditing using Prior Knowledgecited by 6
  12. 2023When Fairness Meets Privacy: Exploring Privacy Threats in Fair Binary Classifiers via Membership Inference Attackscited by 5
  13. 2023Biased Attention: Do Vision Transformers Amplify Gender Bias More than Convolutional Neural Networks?cited by 4
  14. 2023A Theoretical Approach to Characterize the Accuracy-Fairness Trade-off Pareto Frontiercited by 4
  15. 2023The Limits of Fair Medical Imaging AI In The Wildcited by 4
  16. 2024On Biases in a UK Biobank-based Retinal Image Classification Modelcited by 4
  17. 2024PFGuard: A Generative Framework with Privacy and Fairness Safeguardscited by 3
  18. 2024Synthetic Data Generation for Intersectional Fairness by Leveraging Hierarchical Group Structurecited by 2
  19. 2024A Self-Supervised Learning Pipeline for Demographically Fair Facial Attribute Classificationcited by 2
  20. 2024Open Challenges on Fairness of Artificial Intelligence in Medical Imaging Applicationscited by 2
  21. 2025Fairness of Deep Ensembles: On the interplay between per-group task difficulty and under-representationcited by 2
  22. 2025Do Fairness Interventions Come at the Cost of Privacy: Evaluations for Binary Classifierscited by 2
  23. 2026Towards Fairness under Label Bias in Image Segmentation: Impact, Measurement and Mitigationcited by 2
  24. 2022A Differentiable Distance Approximation for Fairer Image Classificationcited by 1
  25. 2022Mitigating Urban-Rural Disparities in Contrastive Representation Learning with Satellite Imagerycited by 1
  26. 2023Linking convolutional kernel size to generalization bias in face analysis CNNscited by 1
  27. 2023Fair Without Leveling Down: A New Intersectional Fairness Definitioncited by 1
  28. 2023HARD: Hard Augmentations for Robust Distillationcited by 1
  29. 2023GELDA: A generative language annotation framework to reveal visual biases in datasetscited by 1
  30. 2025A Multi-Objective Evaluation Framework for Analyzing Utility-Fairness Trade-Offs in Machine Learning Systemscited by 1
  31. 2025Subgroups Matter for Robust Bias Mitigationcited by 1
  32. 2025A Primer on Causal and Statistical Dataset Biases for Fair and Robust Image Analysiscited by 1
  33. 2025OxEnsemble: Fair Ensembles for Low-Data Classificationcited by 1
  34. 2026Stride-Net: Fairness-Aware Disentangled Representation Learning for Chest X-Ray Diagnosiscited by 1
  35. 2023Fairness meets Cross-Domain Learning: a new perspective on Models and Metricscited by 0
  36. 2024On Demographic Group Fairness Guarantees in Deep Learningcited by 0
  37. 2025SWiFT: Soft-Mask Weight Fine-tuning for Bias Mitigationcited by 0
  38. 2025BM-CL: Bias Mitigation through the lens of Continual Learningcited by 0

50 most cited of 53 All 53 (JSON)