Description
AugMix mixes augmented images through linear interpolations. Consequently it is like Mixup but instead mixes augmented versions of the same image.
Papers Using This Method
MediAug: Exploring Visual Augmentation in Medical Imaging2025-04-26Self-supervised Transformation Learning for Equivariant Representations2025-01-15FedCCRL: Federated Domain Generalization with Cross-Client Representation Learning2024-10-15Dynamic Batch Norm Statistics Update for Natural Robustness2023-10-31Navigating Noise: A Study of How Noise Influences Generalisation and Calibration of Neural Networks2023-06-30DiffAug: A Diffuse-and-Denoise Augmentation for Training Robust Classifiers2023-06-15Impact of Light and Shadow on Robustness of Deep Neural Networks2023-05-23What Are Effective Labels for Augmented Data? Improving Calibration and Robustness with AutoLabel2023-02-22Fourier-Based Augmentations for Improved Robustness and Uncertainty Calibration2022-02-24How to augment your ViTs? Consistency loss and StyleAug, a random style transfer augmentation2021-12-16Benchmarks for Corruption Invariant Person Re-identification2021-11-01AugMax: Adversarial Composition of Random Augmentations for Robust Training2021-10-26BulletTrain: Accelerating Robust Neural Network Training via Boundary Example Mining2021-09-29Defending Against Image Corruptions Through Adversarial Augmentations2021-04-02On the effectiveness of adversarial training against common corruptions2021-03-03What are effective labels for augmented data? Improving robustness with AutoLabel2021-01-01A Deep Learning Approach for Diabetic Retinopathy detection using Transfer Learning2021-01-01StackMix: A complementary Mix algorithm2020-11-25Improving robustness against common corruptions by covariate shift adaptation2020-06-30AugMix: A Simple Data Processing Method to Improve Robustness and Uncertainty2019-12-05