[1909.13719] RandAugment: Practical automated data augmentation with a reduced search spaceopen searchopen navigation menucontact arXivsubscribe to arXiv mailings

Recent work has shown that data augmentation has the potential to significantly improve the generalization of deep learning models. Recently, automated augmentation strategies have led to state-of-the-art results in image classification and object detection. While these strategies were optimized for improving validation accuracy, they also led to state-of-the-art results in semi-supervised learning and improved robustness to common corruptions of images. An obstacle to a large-scale adoption of these methods is a separate search phase which increases the training complexity and may substantially increase the computational cost. Additionally, due to the separate search phase, these approaches are unable to adjust the regularization strength based on model or dataset size. Automated augmentation policies are often found by training small models on small datasets and subsequently applied to train larger models. In this work, we remove both of these obstacles. RandAugment has a significant

1 mentions: @AkiraTOSEI
Date: 2020/06/29 09:51

Referring Tweets

@AkiraTOSEI Noisy student自体にRandAug(t.co/gjF1eefDNz)入ってるし、そりゃデータ拡張に強くなるやろ、って感じはする。

Bookmark Comments

Related Entries

Read more 190602 benchmarking neural network robustness to common corruptions a…
0 users, 1 mentions 2019/11/19 15:50
Read more FixMatch ,Semi-Supervised Learning method, that can be learned even if there is only one labeled dat...
0 users, 1 mentions 2020/01/26 07:10
Read more [2001.08764] Fine-Tuning a Transformer-Based Language Model to Avoid Generating Non-Normative Textco...
0 users, 3 mentions 2020/02/04 12:54
Read more [1907.13052] GENESIS: Generative Scene Inference and Sampling with Object-Centric Latent Representat...
0 users, 2 mentions 2020/02/25 14:21
Read more [2005.01234] One-Shot Image Classification by Learning to Restore Prototypesopen searchopen navigati...
0 users, 2 mentions 2020/06/02 11:21