mlsanityData to 14 Oct 2026

Towards Fairer Datasets: Filtering and Balancing the Distribution of the People Subtree in the ImageNet Hierarchy

arXiv 1912.07726 ↗First posted 16 Dec 2019cs.CVcited by 156

Not on the map. Why?

Abstract (arXiv)

Computer vision technology is being used by many but remains representative of only a few. People have reported misbehavior of computer vision models, including offensive prediction results and lower performance for underrepresented groups. Current computer vision models are typically developed using datasets consisting of manually annotated images or videos; the data and label distributions in these datasets are critical to the models' behavior. In this paper, we examine ImageNet, a large-scale ontology of images that has spurred the development of many modern computer vision methods. We consider three key factors within the "person" subtree of ImageNet that may lead to problematic behavior in downstream computer vision technology: (1) the stagnant concept vocabulary of WordNet, (2) the attempt at exhaustive illustration of all categories with images, and (3) the inequality of representation in the images within concepts. We seek to illuminate the root causes of these concerns and take the first steps to mitigate them constructively.

Key facts. Cited by 156 papers in the map (data through 14 Oct 2026). First posted on arXiv 16 Dec 2019.

Builds on

No reference list yet.

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2021: 37; 2022: 39; 2023: 37; 2024: 23; 2025: 12; 2026: 8.

156 papers.

Sort:

By area:

  1. 2023Segment Anythingcited by 8,234
  2. 2023DataComp: In search of the next generation of multimodal datasetscited by 644
  3. 2022Beyond neural scaling laws: beating power law scaling via data pruningcited by 528
  4. 2021Retiring Adult: New Datasets for Fair Machine Learningcited by 449
  5. 2021Intriguing Properties of Vision Transformerscited by 446
  6. 2021SWAD: Domain Generalization by Seeking Flat Minimacited by 434
  7. 2022Kubric: A scalable dataset generatorcited by 381
  8. 2021Visually Grounded Reasoning across Languages and Culturescited by 189
  9. 2022Can Machines Help Us Answering Question 16 in Datasheets, and In Turn Reflecting on Inappropriate Content?cited by 155
  10. 2023Fair Diffusion: Instructing Text-to-Image Generation Models on Fairnesscited by 151
  11. 2022Domain Generalization by Mutual-Information Regularization with Pre-trained Modelscited by 133
  12. 2023Open-domain Visual Entity Recognition: Towards Recognizing Millions of Wikipedia Entitiescited by 113
38 more
  1. 2021Understanding and Evaluating Racial Biases in Image Captioningcited by 108
  2. 2021Pre-training without Natural Imagescited by 100
  3. 2021CLOOB: Modern Hopfield Networks with InfoLOOB Outperform CLIPcited by 89
  4. 2021A Survey on Bias in Visual Datasetscited by 84
  5. 2021A Study of Face Obfuscation in ImageNetcited by 83
  6. 2022Distilling Model Failures as Directions in Latent Spacecited by 82
  7. 2021Towards Measuring Fairness in AI: the Casual Conversations Datasetcited by 80
  8. 2022Vision Models Are More Robust And Fair When Pretrained On Uncurated Images Without Supervisioncited by 80
  9. 2022Learning Audio-Video Modalities from Image Captionscited by 80
  10. 2021Self-Supervised Learning with Kernel Dependence Maximizationcited by 77
  11. 2022Multi-Modal Knowledge Graph Construction and Application: A Surveycited by 74
  12. 2021On Improving Adversarial Transferability of Vision Transformerscited by 70
  13. 2023Uncurated Image-Text Datasets: Shedding Light on Demographic Biascited by 52
  14. 2021PASS: An ImageNet replacement for self-supervised pretraining without humanscited by 51
  15. 2023GeoDE: a Geographically Diverse Evaluation Dataset for Object Recognitioncited by 50
  16. 2024A Decade's Battle on Dataset Bias: Are We There Yet?cited by 47
  17. 2021A Step Toward More Inclusive People Annotations for Fairnesscited by 46
  18. 2022Gender Artifacts in Visual Datasetscited by 38
  19. 2022Benchmarking Omni-Vision Representation through the Lens of Visual Realmscited by 38
  20. 2022Quantifying Societal Bias Amplification in Image Captioningcited by 37
  21. 2022Measuring Representational Harms in Image Captioningcited by 36
  22. 2021Trivial or impossible -- dichotomous data difficulty masks model differences (on ImageNet and beyond)cited by 35
  23. 2021Can Vision Transformers Learn without Natural Images?cited by 32
  24. 2021Representation Matters: Assessing the Importance of Subgroup Allocations in Training Datacited by 31
  25. 2022FairGRAPE: Fairness-aware GRAdient Pruning mEthod for Face Attribute Classificationcited by 30
  26. 2023Balancing the Picture: Debiasing Vision-Language Datasets with Synthetic Contrast Setscited by 27
  27. 2024ViSAGe: A Global-Scale Analysis of Visual Stereotypes in Text-to-Image Generationcited by 27
  28. 2021A Sociotechnical View of Algorithmic Fairnesscited by 24
  29. 2022Fairness Indicators for Systematic Assessments of Visual Feature Extractorscited by 24
  30. 2022Bugs in the Data: How ImageNet Misrepresents Biodiversitycited by 24
  31. 2022On the Strong Correlation Between Model Invariance and Generalizationcited by 23
  32. 2022A Dense Material Segmentation Dataset for Indoor and Outdoor Scene Parsingcited by 23
  33. 2023ImageNet-OOD: Deciphering Modern Out-of-Distribution Detection Algorithmscited by 22
  34. 2022Bamboo: Building Mega-Scale Vision Dataset Continually with Human-Machine Synergycited by 21
  35. 2022Pre-training Vision Transformers with Formula-driven Supervised Learningcited by 21
  36. 2022Men Also Do Laundry: Multi-Attribute Bias Amplificationcited by 21
  37. 2024Real Risks of Fake Data: Synthetic Data, Diversity-Washing and Consent Circumventioncited by 21
  38. 2023Visual Atoms: Pre-training Vision Transformers with Sinusoidal Wavescited by 20

50 most cited of 156 All 156 (JSON)