mlsanityData to 7 Oct 2026

CutMix: Regularization Strategy to Train Strong Classifiers with Localizable Features

arXiv 1905.04899 ↗First posted 13 May 2019cs.CV cs.LGcited by 2,875

Not on the map. Why?

Cited by 92 in Diffusion models, but placed by its own abstract, which didn’t fit.

Abstract (arXiv)

Regional dropout strategies have been proposed to enhance the performance of convolutional neural network classifiers. They have proved to be effective for guiding the model to attend on less discriminative parts of objects (e.g. leg as opposed to head of a person), thereby letting the network generalize better and have better object localization capabilities. On the other hand, current methods for regional dropout remove informative pixels on training images by overlaying a patch of either black pixels or random noise. Such removal is not desirable because it leads to information loss and inefficiency during training. We therefore propose the CutMix augmentation strategy: patches are cut and pasted among training images where the ground truth labels are also mixed proportionally to the area of the patches. By making efficient use of training pixels and retaining the regularization effect of regional dropout, CutMix consistently outperforms the state-of-the-art augmentation strategies on CIFAR and ImageNet classification tasks, as well as on the ImageNet weakly-supervised localization task. Moreover, unlike previous augmentation methods, our CutMix-trained ImageNet classifier, when used as a pretrained model, results in consistent performance gains in Pascal detection and MS-COCO image captioning benchmarks. We also show that CutMix improves the model robustness against input corruptions and its out-of-distribution detection performances. Source code and pretrained models are available at https://github.com/clovaai/CutMix-PyTorch .

Key facts. Cited by 2,875 papers in the map (data through 7 Oct 2026). First posted on arXiv 13 May 2019.

Builds on

No reference list yet.

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2021: 453; 2022: 635; 2023: 630; 2024: 526; 2025: 397; 2026: 234.

2,875 papers.

Sort:

By area:

  1. 2021Swin Transformer: Hierarchical Vision Transformer using Shifted Windowscited by 12,082
  2. 2021Masked Autoencoders Are Scalable Vision Learnerscited by 7,339
  3. 2022A ConvNet for the 2020scited by 3,545
  4. 2021Pyramid Vision Transformer: A Versatile Backbone for Dense Prediction without Convolutionscited by 1,960
  5. 2022Model soups: averaging weights of multiple fine-tuned models improves accuracy without increasing inference timecited by 1,366
  6. 2024Depth Anything: Unleashing the Power of Large-Scale Unlabeled Datacited by 1,346
  7. 2022VideoMAE: Masked Autoencoders are Data-Efficient Learners for Self-Supervised Video Pre-Trainingcited by 1,305
  8. 2021ViLT: Vision-and-Language Transformer Without Convolution or Region Supervisioncited by 1,224
  9. 2021Swin Transformer V2: Scaling Up Capacity and Resolutioncited by 1,215
  10. 2021SimMIM: A Simple Framework for Masked Image Modelingcited by 1,147
  11. 2021Tokens-to-Token ViT: Training Vision Transformers from Scratch on ImageNetcited by 1,096
  12. 2021Transformers in Vision: A Surveycited by 1,009
38 more
  1. 2023Self-Supervised Learning from Images with a Joint-Embedding Predictive Architecturecited by 982
  2. 2021Multiscale Vision Transformerscited by 896
  3. 2021CrossViT: Cross-Attention Multi-Scale Vision Transformer for Image Classificationcited by 806
  4. 2021Generalized Out-of-Distribution Detection: A Surveycited by 796
  5. 2022OFA: Unifying Architectures, Tasks, and Modalities Through a Simple Sequence-to-Sequence Learning Frameworkcited by 742
  6. 2022AdaptFormer: Adapting Vision Transformers for Scalable Visual Recognitioncited by 714
  7. 2023ConvNeXt V2: Co-designing and Scaling ConvNets with Masked Autoencoderscited by 703
  8. 2022EVA: Exploring the Limits of Masked Visual Representation Learning at Scalecited by 685
  9. 2021Point-BERT: Pre-training 3D Point Cloud Transformers with Masked Point Modelingcited by 647
  10. 2021Transformer in Transformercited by 632
  11. 2021Perceiver IO: A General Architecture for Structured Inputs & Outputscited by 616
  12. 2021Masked Feature Prediction for Self-Supervised Visual Pre-Trainingcited by 573
  13. 2021Domain Generalization with MixStylecited by 563
  14. 2021CutPaste: Self-Supervised Learning for Anomaly Detection and Localizationcited by 529
  15. 2021CSWin Transformer: A General Vision Transformer Backbone with Cross-Shaped Windowscited by 526
  16. 2021MetaFormer Is Actually What You Need for Visioncited by 523
  17. 2021MViTv2: Improved Multiscale Vision Transformers for Classification and Detectioncited by 516
  18. 2023VideoMAE V2: Scaling Video Masked Autoencoders with Dual Maskingcited by 516
  19. 2021Deep Neural Networks and Tabular Data: A Surveycited by 489
  20. 2021Semi-Supervised Semantic Segmentation with Cross Pseudo Supervisioncited by 462
  21. 2021SWAD: Domain Generalization by Seeking Flat Minimacited by 431
  22. 2022Masked Autoencoders As Spatiotemporal Learnerscited by 428
  23. 2022Diffusion Art or Digital Forgery? Investigating Data Replication in Diffusion Modelscited by 418
  24. 2022DeiT III: Revenge of the ViTcited by 417
  25. 2021Deep Long-Tailed Learning: A Surveycited by 416
  26. 2021A Survey of Data Augmentation Approaches for NLPcited by 408
  27. 2021Conditional Positional Encodings for Vision Transformerscited by 402
  28. 2021Early Convolutions Help Transformers See Bettercited by 395
  29. 2023Hyena Hierarchy: Towards Larger Convolutional Language Modelscited by 388
  30. 2023EVA-02: A Visual Representation for Neon Genesiscited by 377
  31. 2021ResNet strikes back: An improved training procedure in timmcited by 365
  32. 2021Dataset Condensation with Distribution Matchingcited by 362
  33. 2021Rethinking Spatial Dimensions of Vision Transformerscited by 361
  34. 2021High-Performance Large-Scale Image Recognition Without Normalizationcited by 359
  35. 2021ResMLP: Feedforward networks for image classification with data-efficient trainingcited by 338
  36. 2023SVDiff: Compact Parameter Space for Diffusion Fine-Tuningcited by 338
  37. 2022Scaling Language-Image Pre-training via Maskingcited by 317
  38. 2021SAINT: Improved Neural Networks for Tabular Data via Row Attention and Contrastive Pre-Trainingcited by 310

50 most cited of 2,875 All 2,875 (JSON)