Ctrl-A: Control-Driven Online Data Augmentation

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Christensen, Jesper B., Bench, Ciaran, Thomas, Spencer A., Aslan, Hüsnü, Balslev-Harder, David, Smith, Nadia A. S., Manzin, Alessandra
Format: Preprint
Published: 2026
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866910065483579392
author Christensen, Jesper B.
Bench, Ciaran
Thomas, Spencer A.
Aslan, Hüsnü
Balslev-Harder, David
Smith, Nadia A. S.
Manzin, Alessandra
author_facet Christensen, Jesper B.
Bench, Ciaran
Thomas, Spencer A.
Aslan, Hüsnü
Balslev-Harder, David
Smith, Nadia A. S.
Manzin, Alessandra
contents We introduce ControlAugment (Ctrl-A), an automated data augmentation algorithm for image-vision tasks, which incorporates principles from control theory for online adjustment of augmentation strength distributions during model training. Ctrl-A eliminates the need for initialization of individual augmentation strengths. Instead, augmentation strength distributions are dynamically, and individually, adapted during training based on a control-loop architecture and what we define as relative operation response curves. Using an operation-dependent update procedure provides Ctrl-A with the potential to suppress augmentation styles that negatively impact model performance, alleviating the need for manually engineering augmentation policies for new image-vision tasks. Experiments on the CIFAR-10, CIFAR-100, and SVHN-core benchmark datasets using the common WideResNet-28-10 architecture demonstrate that Ctrl-A is highly competitive with existing state-of-the-art data augmentation strategies.
format Preprint
id arxiv_https___arxiv_org_abs_2603_21819
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Ctrl-A: Control-Driven Online Data Augmentation
Christensen, Jesper B.
Bench, Ciaran
Thomas, Spencer A.
Aslan, Hüsnü
Balslev-Harder, David
Smith, Nadia A. S.
Manzin, Alessandra
Computer Vision and Pattern Recognition
Artificial Intelligence
Machine Learning
Systems and Control
We introduce ControlAugment (Ctrl-A), an automated data augmentation algorithm for image-vision tasks, which incorporates principles from control theory for online adjustment of augmentation strength distributions during model training. Ctrl-A eliminates the need for initialization of individual augmentation strengths. Instead, augmentation strength distributions are dynamically, and individually, adapted during training based on a control-loop architecture and what we define as relative operation response curves. Using an operation-dependent update procedure provides Ctrl-A with the potential to suppress augmentation styles that negatively impact model performance, alleviating the need for manually engineering augmentation policies for new image-vision tasks. Experiments on the CIFAR-10, CIFAR-100, and SVHN-core benchmark datasets using the common WideResNet-28-10 architecture demonstrate that Ctrl-A is highly competitive with existing state-of-the-art data augmentation strategies.
title Ctrl-A: Control-Driven Online Data Augmentation
topic Computer Vision and Pattern Recognition
Artificial Intelligence
Machine Learning
Systems and Control
url https://arxiv.org/abs/2603.21819