mlsanityData to 14 Oct 2026

Explaining Bias in Deep Face Recognition via Image Characteristics

arXiv 2208.11099 ↗First posted 23 Aug 2022cs.CVcited by 6

Not on the map. Why?

Abstract (arXiv)

In this paper, we propose a novel explanatory framework aimed to provide a better understanding of how face recognition models perform as the underlying data characteristics (protected attributes: gender, ethnicity, age; non-protected attributes: facial hair, makeup, accessories, face orientation and occlusion, image distortion, emotions) on which they are tested change. With our framework, we evaluate ten state-of-the-art face recognition models, comparing their fairness in terms of security and usability on two data sets, involving six groups based on gender and ethnicity. We then analyze the impact of image characteristics on models performance. Our results show that trends appearing in a single-attribute analysis disappear or reverse when multi-attribute groups are considered, and that performance disparities are also related to non-protected attributes. Source code: https://cutt.ly/2XwRLiA.

Key facts. Cited by 6 papers in the map (data through 14 Oct 2026). Builds on 26 of its 42 references that are in the map. First posted on arXiv 23 Aug 2022.

Builds on

26 of 42 references.

References · 26

  1. 2015Deep Residual Learning for Image Recognitioncited by 52,096
  2. 2019A Survey on Bias and Fairness in Machine Learningcited by 1,917
  3. 2019Deep High-Resolution Representation Learning for Visual Recognitioncited by 1,420
  4. 2017VGGFace2: A dataset for recognising faces across pose and agecited by 914
  5. 2016Inherent Trade-Offs in the Fair Determination of Risk Scorescited by 586
  6. 2021RepVGG: Making VGG-style ConvNets Great Againcited by 522
  7. 2020ResNeSt: Split-Attention Networkscited by 445
  8. 2017Residual Attention Network for Image Classificationcited by 386
  9. 2018Deep Face Recognition: A Surveycited by 276
  10. 2021A Comprehensive Study on Face Recognition Biases Beyond Demographicscited by 86
  11. 2020Mitigating Face Recognition Bias via Group Adaptive Classifiercited by 67
  12. 2022Meta Balanced Network for Fair Face Recognitioncited by 53
14 more
  1. 2021FaceX-Zoo: A PyTorch Toolbox for Face Recognitioncited by 49
  2. 2020Analysis of Gender Inequality In Face Recognition Accuracycited by 47
  3. 2020Face Quality Estimation and Its Correlation to Demographic and Non-Demographic Bias in Face Recognitioncited by 31
  4. 2020InsideBias: Measuring Bias in Deep Networks and Application to Face Gender Biometricscited by 31
  5. 2020FairFace Challenge at ECCV 2020: Analyzing Bias in Face Recognitioncited by 26
  6. 2019Deep Learning for Face Recognition: Pride or Prejudiced?cited by 25
  7. 2020Is Face Recognition Sexist? No, Gendered Hairstyles and Biology Arecited by 23
  8. 2021Rethinking Common Assumptions to Mitigate Racial Bias in Face Recognition Datasetscited by 22
  9. 2021FairCal: Fairness Calibration for Face Verificationcited by 17
  10. 2021Unravelling the Effect of Image Distortions for Biased Prediction of Pre-trained Face Recognition Modelscited by 11
  11. 2021Robustness Disparities in Commercial Face Detectioncited by 11
  12. 2020NPCFace: Negative-Positive Collaborative Training for Large-scale Face Recognitioncited by 9
  13. 2021Fairness Properties of Face Recognition and Obfuscation Systemscited by 9
  14. 2021Face Verification with Challenging Imposters and Diversified Demographicscited by 5

Built on by

Cited by year

Citing papers per year: 2022: 2; 2023: 1; 2024: 1; 2025: 2.

6 papers.

  1. 2025Review of Demographic Fairness in Face Recognitioncited by 7
  2. 2022A survey of Identification and mitigation of Machine Learning algorithmic biases in Image Analysiscited by 3
  3. 2024The Impact of Balancing Real and Synthetic Data on Accuracy and Fairness in Face Recognitioncited by 3
  4. 2022The More Secure, The Less Equally Usable: Gender and Ethnicity (Un)fairness of Deep Face Recognition along Security Thresholdscited by 2
  5. 2023(Un)fair Exposure in Deep Face Rankings at a Distancecited by 1
  6. 2025DemoBias: An Empirical Study to Trace Demographic Biases in Vision Foundation Modelscited by 1