Randaugment: Practical automated data augmentation with a reduced search space
RandAugment: практическое автоматизированное расширение данных с сокращённым пространством поиска
2020-06-01
SCID: 54.1/tadm24dd
Discuss with AI
RandAugmentautomated data augmentationimage classificationobject detectionreduced search space
Figures from the paper
Abstract (AI)
Recent work on automated augmentation strategies has led to state-of-the-art results in image classification and object detection. An obstacle to a large-scale adoption of these methods is that they require a separate and expensive search phase. A common way to overcome the expense of the search phase was to use a smaller proxy task. However, it was not clear if the optimized hyperparameters found on the proxy task are also optimal for the actual task. In this work, we rethink the process of designing automated augmentation strategies. We find that while previous work required a search for both magnitude and probability of each operation independently, it is sufficient to only search for a single distortion magnitude that jointly controls all operations. We hence propose a simplified search space that vastly reduces the computational expense of automated augmentation, and permits the removal of a separate proxy task. Despite the simplifications, our method achieves equal or better performance over previous automated augmentation strategies on on CIFAR-10/100, SVHN, ImageNet and COCO datasets. EfficientNet-B7, we achieve 85.0% accuracy, a 1.0% increase over baseline augmentation, a 0.6% improvement over AutoAugment on the ImageNet dataset. With EfficientNet-B8, we achieve 85.4% accuracy on ImageNet, which matches a previous result that used 3.5B extra images. On object detection, the same method as classification leads to 1.0-1.3% improvement over baseline augmentation. Code will be made available online.
Key Findings
1
RandAugment matches or outperforms previous automated augmentation strategies on CIFAR-10/100, SVHN, ImageNet, and COCO.
2
RandAugment reduces automated augmentation search to a single distortion magnitude jointly controlling all operations, eliminating independent magnitude and probability searches.
3
The simplified search space substantially lowers computational cost and removes the need for a separate proxy task.
4
With EfficientNet-B7, RandAugment achieves 85.0% ImageNet accuracy, improving 1.0 percentage point over baseline augmentation and 0.6 points over AutoAugment.
5
With EfficientNet-B8, RandAugment reaches 85.4% ImageNet accuracy, matching a result that used 3.5 billion additional images; object detection improves 1.0–1.3 percentage points over baseline augmentation.
Research Object
automated image augmentation strategies for image classification and object detection
Research Subject
the effectiveness and computational efficiency of a reduced search space using a single joint distortion magnitude instead of independently searching operation magnitudes and probabilities
Publication Details
Publication Date
2020-06-01
Journal
Publisher
ISSN
Cited by
5030
Open access PDF
Access Type
Author Information
Download PDF
Subscribe to digest
References available in scid.ai8
Deep Residual Learning for Image Recognition2016
ImageNet classification with deep convolutional neural networks2017
ImageNet: A large-scale hierarchical image database2009
Rethinking the Inception Architecture for Computer Vision2016
Learning Multiple Layers of Features from Tiny Images2024
Intriguing properties of neural networks2013
Deep Neural Networks for Acoustic Modeling in Speech Recognition: The Shared Views of Four Research Groups2012
mixup: Beyond Empirical Risk Minimization2017
Cited by13
Swin Transformer: Hierarchical Vision Transformer using Shifted Windows2021
Swin Transformer V2: Scaling Up Capacity and Resolution2022
Run, Don't Walk: Chasing Higher FLOPS for Faster Neural Networks2023
Video Swin Transformer2022
CrossViT: Cross-Attention Multi-Scale Vision Transformer for Image Classification2021
Text Data Augmentation for Deep Learning2021
CSWin Transformer: A General Vision Transformer Backbone with Cross-Shaped Windows2022
Vision Transformer with Deformable Attention2022
Efficient Deep Learning: A Survey on Making Deep Learning Models Smaller, Faster, and Better2023
A survey on data‐efficient algorithms in big data era2021
Transformer Neural Network for Weed and Crop Classification of High Resolution UAV Images2022
Automated data processing and feature engineering for deep learning and big data applications: A survey2024
Graph Neural Networks for Pressure Estimation in Water Distribution Systems2024