mlsanityData to 14 Oct 2026

Residual Relaxation for Multi-view Representation Learning

arXiv 2110.15348 ↗First posted 28 Oct 2021cs.LG cs.CVcited by 30

Not on the map. Why?

Abstract (arXiv)

Multi-view methods learn representations by aligning multiple views of the same image and their performance largely depends on the choice of data augmentation. In this paper, we notice that some other useful augmentations, such as image rotation, are harmful for multi-view methods because they cause a semantic shift that is too large to be aligned well. This observation motivates us to relax the exact alignment objective to better cultivate stronger augmentations. Taking image rotation as a case study, we develop a generic approach, Pretext-aware Residual Relaxation (Prelax), that relaxes the exact alignment by allowing an adaptive residual vector between different views and encoding the semantic shift through pretext-aware learning. Extensive experiments on different backbones show that our method can not only improve multi-view methods with existing augmentations, but also benefit from stronger image augmentations like rotation.

Key facts. Cited by 30 papers in the map (data through 14 Oct 2026). Builds on 24 of its 35 references that are in the map. First posted on arXiv 28 Oct 2021.

Builds on

24 of 35 references.

References · 24

  1. 2015Deep Residual Learning for Image Recognitioncited by 52,096
  2. 2014Microsoft COCO: Common Objects in Contextcited by 17,608
  3. 2020A Simple Framework for Contrastive Learning of Visual Representationscited by 12,049
  4. 2018Representation Learning with Contrastive Predictive Codingcited by 7,777
  5. 2019Momentum Contrast for Unsupervised Visual Representation Learningcited by 7,513
  6. 2017Mask R-CNNcited by 7,277
  7. 2016SGDR: Stochastic Gradient Descent with Warm Restartscited by 5,045
  8. 2016Feature Pyramid Networks for Object Detectioncited by 5,036
  9. 2020Bootstrap your own latent: A new approach to self-supervised Learningcited by 4,750
  10. 2020Exploring Simple Siamese Representation Learningcited by 2,866
  11. 2020Improved Baselines with Momentum Contrastive Learningcited by 2,287
  12. 2018Unsupervised Representation Learning by Predicting Image Rotationscited by 1,591
12 more
  1. 2016Unsupervised Learning of Visual Representations by Solving Jigsaw Puzzlescited by 1,382
  2. 2020Big Self-Supervised Models are Strong Semi-Supervised Learnerscited by 1,380
  3. 2019Contrastive Multiview Codingcited by 1,301
  4. 2018Learning deep representations by mutual information estimation and maximizationcited by 1,273
  5. 2016Colorful Image Colorizationcited by 1,114
  6. 2015Unsupervised Visual Representation Learning by Context Predictioncited by 1,066
  7. 2020What Makes for Good Views for Contrastive Learning?cited by 846
  8. 2019Self-Supervised Learning of Pretext-Invariant Representationscited by 808
  9. 2017Focal Loss for Dense Object Detectioncited by 752
  10. 2019Learning Representations by Maximizing Mutual Information Across Viewscited by 734
  11. 2020What Should Not Be Contrastive in Contrastive Learningcited by 226
  12. 2016A Survey of Multi-View Representation Learningcited by 107

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2022: 7; 2023: 11; 2024: 6; 2025: 4; 2026: 2.

30 papers.

Sort:

By area:

  1. 2022Chaos is a Ladder: A New Theoretical Understanding of Contrastive Learning via Augmentation Overlapcited by 105
  2. 2022How Mask Matters: Towards Theoretical Understandings of Masked Autoencoderscited by 86
  3. 2023Towards a Unified Theoretical Understanding of Non-contrastive Learning via Rank Differential Mechanismcited by 33
  4. 2023On the Generalization of Multi-modal Contrastive Learningcited by 30
  5. 2024A Canonicalization Perspective on Invariant and Equivariant Learningcited by 27
  6. 2024Do Generated Data Always Help Contrastive Learning?cited by 25
  7. 2023A Message Passing Perspective on Learning Dynamics of Contrastive Learningcited by 22
  8. 2023Rethinking the Effect of Data Augmentation in Adversarial Contrastive Learningcited by 20
  9. 2024Non-negative Contrastive Learningcited by 19
  10. 2023Rethinking Weak Supervision in Helping Contrastive Learningcited by 18
  11. 2023Adversarial Examples Are Not Real Featurescited by 16
  12. 2023A Novel Approach for Effective Multi-View Clustering with Information-Theoretic Perspectivecited by 14
18 more
  1. 2023Identifiable Contrastive Learning with Automatic Feature Importance Discoverycited by 14
  2. 2025Projection Head is Secretly an Information Bottleneckcited by 11
  3. 2022Decoupled Self-supervised Learning for Non-Homophilous Graphscited by 10
  4. 2023Architecture Matters: Uncovering Implicit Mechanisms in Graph Contrastive Learningcited by 7
  5. 2024On the Role of Discrete Tokenization in Visual Representation Learningcited by 7
  6. 2024Understanding the Role of Equivariance in Self-supervised Learningcited by 7
  7. 2022Feature Disentanglement Learning with Switching and Aggregation for Video-based Person Re-Identificationcited by 5
  8. 2024Towards Unsupervised Domain Bridging via Image Degradation in Semantic Segmentationcited by 5
  9. 2025An Augmentation Overlap Theory of Contrastive Learningcited by 5
  10. 2023Self-Supervised Learning for Place Representation Generalization across Appearance Changescited by 4
  11. 2022Rethinking Rotation in Self-Supervised Contrastive Learning: Adaptive Positive or Negative Data Augmentationcited by 3
  12. 2022Homomorphic Self-Supervised Learningcited by 3
  13. 2022Mitigating Forgetting in Online Continual Learning via Contrasting Semantically Distinct Augmentationscited by 2
  14. 2025An Inclusive Theoretical Framework of Robust Supervised Contrastive Loss against Label Noisecited by 2
  15. 2025Difficult Examples Hurt Unsupervised Contrastive Learning: A Theoretical Perspectivecited by 2
  16. 2023CIPER: Combining Invariant and Equivariant Representations Using Contrastive and Predictive Learningcited by 1
  17. 2026Adaptive Augmentation-Aware Latent Learning for Robust LiDAR Semantic Segmentationcited by 0
  18. 2026A Theoretical Framework for Masked Pretraining (MPT)cited by 0