mlsanityData to 7 Oct 2026

CLAD: A Contrastive Learning based Approach for Background Debiasing

arXiv 2210.02748 ↗First posted 6 Oct 2022cs.CV cs.AI cs.LGcited by 6

Not on the map. Why?

Abstract (arXiv)

Convolutional neural networks (CNNs) have achieved superhuman performance in multiple vision tasks, especially image classification. However, unlike humans, CNNs leverage spurious features, such as background information to make decisions. This tendency creates different problems in terms of robustness or weak generalization performance. Through our work, we introduce a contrastive learning-based approach (CLAD) to mitigate the background bias in CNNs. CLAD encourages semantic focus on object foregrounds and penalizes learning features from irrelavant backgrounds. Our method also introduces an efficient way of sampling negative samples. We achieve state-of-the-art results on the Background Challenge dataset, outperforming the previous benchmark with a margin of 4.1%. Our paper shows how CLAD serves as a proof of concept for debiasing of spurious features, such as background and texture (in supplementary material).

Key facts. Cited by 6 papers in the map (data through 7 Oct 2026). Builds on 34 of its 48 references that are in the map. First posted on arXiv 6 Oct 2022.

Builds on

34 of 48 references.

References · 34

  1. 2015Deep Residual Learning for Image Recognitioncited by 51,809
  2. 2014Adam: A Method for Stochastic Optimizationcited by 36,960
  3. 2020A Simple Framework for Contrastive Learning of Visual Representationscited by 11,979
  4. 2018Representation Learning with Contrastive Predictive Codingcited by 7,662
  5. 2019Momentum Contrast for Unsupervised Visual Representation Learningcited by 7,482
  6. 2020Bootstrap your own latent: A new approach to self-supervised Learningcited by 4,712
  7. 2014Going Deeper with Convolutionscited by 4,231
  8. 2020Supervised Contrastive Learningcited by 2,803
  9. 2017Arbitrary Style Transfer in Real-time with Adaptive Instance Normalizationcited by 1,998
  10. 2018ImageNet-trained CNNs are biased towards texture; increasing shape bias improves accuracy and robustnesscited by 1,430
  11. 2022DINO: DETR with Improved DeNoising Anchor Boxes for End-to-End Object Detectioncited by 1,306
  12. 2017SmoothGrad: removing noise by adding noisecited by 1,097
22 more
  1. 2019Learning Robust Global Representations by Penalizing Local Predictive Powercited by 1,010
  2. 2018Recognition in Terra Incognitacited by 655
  3. 2020U2-Net: Going Deeper with Nested U-Structure for Salient Object Detectioncited by 410
  4. 2020Noise or Signal: The Role of Image Backgrounds in Object Recognitioncited by 342
  5. 2019A Fourier Perspective on Model Robustness in Computer Visioncited by 288
  6. 2019Approximating CNNs with Bag-of-local-Features models works surprisingly well on ImageNetcited by 260
  7. 2019Inherent Tradeoffs in Learning Fair Representationscited by 144
  8. 2022What to Hide from Your Students: Attention-Guided Masked Image Modelingcited by 117
  9. 2022Contrastive Learning Rivals Masked Image Modeling in Fine-tuning via Feature Distillationcited by 93
  10. 2020Social NCE: Contrastive Learning of Socially-aware Motion Representationscited by 58
  11. 2021Adversarial Robustness Comparison of Vision Transformer and MLP-Mixer to CNNscited by 57
  12. 2022A Comprehensive Study of Image Classification Model Sensitivity to Foregrounds, Backgrounds, and Visual Attributescited by 56
  13. 2020Removing Spurious Features can Hurt Accuracy and Affect Groups Disproportionatelycited by 52
  14. 2021Shape or Texture: Understanding Discriminative Features in CNNscited by 52
  15. 2016Object Recognition with and without Objectscited by 39
  16. 2017Comparing deep neural networks against humans: object recognition when the signal gets weakercited by 34
  17. 2021Object-aware Contrastive Learning for Debiased Scene Representationcited by 34
  18. 2020Hold me tight! Influence of discriminative features on deep network boundariescited by 31
  19. 2021CLCC: Contrastive Learning for Color Constancycited by 29
  20. 2021Robust Representation Learning via Perceptual Similarity Metricscited by 23
  21. 2020Time for a Background Check! Uncovering the impact of Background Features on Deep Neural Networkscited by 5
  22. 2022Empirical Advocacy of Bio-inspired Models for Robust Image Recognitioncited by 3

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2023: 1; 2024: 2; 2025: 2; 2026: 1.

6 papers.

  1. 2024Separating common from salient patterns with Contrastive Representation Learningcited by 6
  2. 2025Efficient Calisthenics Skills Classification through Foreground Instance Selection and Depth Estimationcited by 1
  3. 2025Robust Context-Aware Object Recognitioncited by 1
  4. 2023Leveraging Habitat Information for Fine-grained Bird Identificationcited by 0
  5. 2024Bringing the Context Back into Object Recognition, Robustlycited by 0
  6. 2026ContextShift: A Controlled Benchmark for Context Dependence in Object Detectioncited by 0