Combined Image Data Augmentations diminish the benefits of Adaptive Label Smoothing

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Siedel, Georg, Gupta, Ekagra, Shao, Weijia, Vock, Silvia, Morozov, Andrey
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866912496575578112
author Siedel, Georg
Gupta, Ekagra
Shao, Weijia
Vock, Silvia
Morozov, Andrey
author_facet Siedel, Georg
Gupta, Ekagra
Shao, Weijia
Vock, Silvia
Morozov, Andrey
contents Soft augmentation regularizes the supervised learning process of image classifiers by reducing label confidence of a training sample based on the magnitude of random-crop augmentation applied to it. This paper extends this adaptive label smoothing framework to other types of aggressive augmentations beyond random-crop. Specifically, we demonstrate the effectiveness of the method for random erasing and noise injection data augmentation. Adaptive label smoothing permits stronger regularization via higher-intensity Random Erasing. However, its benefits vanish when applied with a diverse range of image transformations as in the state-of-the-art TrivialAugment method, and excessive label smoothing harms robustness to common corruptions. Our findings suggest that adaptive label smoothing should only be applied when the training data distribution is dominated by a limited, homogeneous set of image transformation types.
format Preprint
id arxiv_https___arxiv_org_abs_2507_16427
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Combined Image Data Augmentations diminish the benefits of Adaptive Label Smoothing
Siedel, Georg
Gupta, Ekagra
Shao, Weijia
Vock, Silvia
Morozov, Andrey
Computer Vision and Pattern Recognition
Machine Learning
Soft augmentation regularizes the supervised learning process of image classifiers by reducing label confidence of a training sample based on the magnitude of random-crop augmentation applied to it. This paper extends this adaptive label smoothing framework to other types of aggressive augmentations beyond random-crop. Specifically, we demonstrate the effectiveness of the method for random erasing and noise injection data augmentation. Adaptive label smoothing permits stronger regularization via higher-intensity Random Erasing. However, its benefits vanish when applied with a diverse range of image transformations as in the state-of-the-art TrivialAugment method, and excessive label smoothing harms robustness to common corruptions. Our findings suggest that adaptive label smoothing should only be applied when the training data distribution is dominated by a limited, homogeneous set of image transformation types.
title Combined Image Data Augmentations diminish the benefits of Adaptive Label Smoothing
topic Computer Vision and Pattern Recognition
Machine Learning
url https://arxiv.org/abs/2507.16427