Fast AutoAugment

Sungbin Lim, Ildoo Kim, Taesup Kim, Chiheon Kim, Sungwoong Kim
5/1/2019
cs.LGcs.CVstat.ML

Official code: kakaobrain/fast-autoaugment (1,604 stars, MIT, PyTorch, last commit 2021-06-16).

Abstract

Data augmentation is an essential technique for improving generalization ability of deep learning models. Recently, AutoAugment has been proposed as an algorithm to automatically search for augmentation policies from a dataset and has significantly enhanced performances on many image recognition tasks. However, its search method requires thousands of GPU hours even for a relatively small dataset. In this paper, we propose an algorithm called Fast AutoAugment that finds effective augmentation policies via a more efficient search strategy based on density matching. In comparison to AutoAugment, the proposed algorithm speeds up the search time by orders of magnitude while achieves comparable performances on image recognition tasks with various models and datasets including CIFAR-10, CIFAR-100, SVHN, and ImageNet.

View on arXivView PDF

Code Implementations(1)

kakaobrain/fast-autoaugmentOfficial100%
PyTorchMIT

Official Implementation of 'Fast AutoAugment' in PyTorch.

1,604195PythonMay 1, 20195 years agoMIT
augmentationautomated-machine-learningautomlcnncomputer-vision+5 more

Similar Papers

AutoAugment: Learning Augmentation Policies from Data

88% match1 repo77.7k2018

LA3: Efficient Label-Aware AutoAugment

87% match1 repo42023

Dynamic Data Augmentation via MCTS for Prostate MRI Segmentation

85% match2023

Adaptive Augmentation Policy Optimization with LLM Feedback

85% match2024

Auto Machine Learning for Medical Image Analysis by Unifying the Search on Data Augmentation and Neural Architecture

84% match2022

Learning Data Augmentation Strategies for Object Detection

83% match1 repo5.3k2019

Cite this paper

@article{lim2019fast,
  title  = {Fast AutoAugment},
  author = {Sungbin Lim and Ildoo Kim and Taesup Kim and Chiheon Kim and Sungwoong Kim},
  year   = {2019},
  eprint = {1905.00397},
  archivePrefix = {arXiv},
  url    = {https://arxiv.org/abs/1905.00397}
}

Add a badge to your README

Shows whether official code is linked for this paper.

Code of Paper badge
[![Code of Paper](https://codeofpaper.com/badge/1905.00397.svg)](https://codeofpaper.com/paper/1905.00397)

Discussion