mlsanityData to 7 Oct 2026

Semantic Equivariant Mixup

arXiv 2308.06451 ↗First posted 12 Aug 2023cs.CV cs.AIcited by 0

Not on the map. Why?

Abstract (arXiv)

Mixup is a well-established data augmentation technique, which can extend the training distribution and regularize the neural networks by creating ''mixed'' samples based on the label-equivariance assumption, i.e., a proportional mixup of the input data results in the corresponding labels being mixed in the same proportion. However, previous mixup variants may fail to exploit the label-independent information in mixed samples during training, which usually contains richer semantic information. To further release the power of mixup, we first improve the previous label-equivariance assumption by the semantic-equivariance assumption, which states that the proportional mixup of the input data should lead to the corresponding representation being mixed in the same proportion. Then a generic mixup regularization at the representation level is proposed, which can further regularize the model with the semantic information in mixed samples. At a high level, the proposed semantic equivariant mixup (sem) encourages the structure of the input data to be preserved in the representation space, i.e., the change of input will result in the obtained representation information changing in the same way. Different from previous mixup variants, which tend to over-focus on the label-related information, the proposed method aims to preserve richer semantic information in the input with semantic-equivariance assumption, thereby improving the robustness of the model against distribution shifts. We conduct extensive empirical studies and qualitative analyzes to demonstrate the effectiveness of our proposed method. The code of the manuscript is in the supplement.

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Builds on 42 of its 57 references that are in the map. First posted on arXiv 12 Aug 2023.

Builds on

42 of 57 references.

References · 42

  1. 2017mixup: Beyond Empirical Risk Minimizationcited by 5,155
  2. 2016Wide Residual Networkscited by 3,057
  3. 2019CutMix: Regularization Strategy to Train Strong Classifiers with Localizable Featurescited by 2,875
  4. 2019Benchmarking Neural Network Robustness to Common Corruptions and Perturbationscited by 2,857
  5. 2018UMAP: Uniform Manifold Approximation and Projection for Dimension Reductioncited by 2,841
  6. 2016Identity Mappings in Deep Residual Networkscited by 1,960
  7. 2020Shortcut Learning in Deep Neural Networkscited by 1,928
  8. 2015LSUN: Construction of a Large-scale Image Dataset using Deep Learning with Humans in the Loopcited by 1,350
  9. 2016Group Equivariant Convolutional Networkscited by 1,154
  10. 2021E(n) Equivariant Graph Neural Networkscited by 951
  11. 2018Tensor field networks: Rotation- and translation-equivariant neural networks for 3D point cloudscited by 735
  12. 2020SE(3)-Transformers: 3D Roto-Translation Equivariant Attention Networkscited by 602
30 more
  1. 2021Domain Generalization with MixStylecited by 563
  2. 2018Spherical CNNscited by 452
  3. 2017A Downsampled Variant of ImageNet as an Alternative to the CIFAR datasetscited by 374
  4. 2020The Pitfalls of Simplicity Bias in Neural Networkscited by 349
  5. 2019On Mixup Training: Improved Calibration and Predictive Uncertainty for Deep Neural Networkscited by 304
  6. 2019Gauge Equivariant Convolutional Networks and the Icosahedral CNNcited by 262
  7. 2020Gradient Starvation: A Learning Proclivity in Neural Networkscited by 262
  8. 2020Puzzle Mix: Exploiting Saliency and Local Statistics for Optimal Mixupcited by 260
  9. 2014Understanding image representations by measuring their equivariance and equivalencecited by 240
  10. 2018A General Theory of Equivariant CNNs on Homogeneous Spacescited by 223
  11. 2022Improving Out-of-Distribution Robustness via Selective Augmentationcited by 202
  12. 2018Do CIFAR-10 Classifiers Generalize to CIFAR-10?cited by 197
  13. 2019Universal Invariant and Equivariant Graph Neural Networkscited by 182
  14. 2020SaliencyMix: A Saliency Guided Data Augmentation Strategy for Better Regularizationcited by 149
  15. 2022G-Mixup: Graph Data Augmentation for Graph Classificationcited by 145
  16. 2021Fair Mixup: Fairness via Interpolationcited by 109
  17. 2016Rotation equivariant vector field networkscited by 98
  18. 2021Equivariant Contrastive Learningcited by 65
  19. 2020Attentive Group Equivariant Convolutional Networkscited by 62
  20. 2021When and How Mixup Improves Calibrationcited by 59
  21. 2021Contrast and Mix: Temporal Contrastive Video Domain Adaptation with Background Mixingcited by 49
  22. 2022UMIX: Improving Importance Weighting for Subpopulation Shift via Uncertainty-Aware Mixupcited by 34
  23. 2021AlignMixup: Improving Representations By Interpolating Aligned Featurescited by 33
  24. 2022A Unified Analysis of Mixed Sample Data Augmentation: A Loss Function Perspectivecited by 27
  25. 2022EquiMod: An Equivariance Module to Improve Self-Supervised Learningcited by 20
  26. 2023The Benefits of Mixup for Feature Learningcited by 19
  27. 2022Provably Learning Diverse Features in Multi-View Data with Midpoint Mixupcited by 14
  28. 2023Over-training with Mixup May Hurt Generalizationcited by 13
  29. 2022Exploring Temporally Dynamic Data Augmentation for Video Recognitioncited by 9
  30. 2023Reweighted Mixup for Subpopulation Shiftcited by 2

Built on by

Not cited yet.