MetaAugment: Sample-Aware Data Augmentation Policy Learning

Authors

  • Fengwei Zhou Huawei Noah's Ark Lab
  • Jiawei Li Huawei Noah's Ark Lab
  • Chuanlong Xie Huawei Noah's Ark Lab
  • Fei Chen Huawei Noah's Ark Lab
  • Lanqing Hong Huawei Noah's Ark Lab
  • Rui Sun Huawei Noah Ark's Lab
  • Zhenguo Li Huawei Noah's Ark Lab

Keywords:

(Deep) Neural Network Algorithms

Abstract

Automated data augmentation has shown superior performance in image recognition. Existing works search for dataset-level augmentation policies without considering individual sample variations, which are likely to be sub-optimal. On the other hand, learning different policies for different samples naively could greatly increase the computing cost. In this paper, we learn a sample-aware data augmentation policy efficiently by formulating it as a sample reweighting problem. Specifically, an augmentation policy network takes a transformation and the corresponding augmented image as inputs, and outputs a weight to adjust the augmented image loss computed by a task network. At training stage, the task network minimizes the weighted losses of augmented training images, while the policy network minimizes the loss of the task network on a validation set via meta-learning. We theoretically prove the convergence of the training procedure and further derive the exact convergence rate. Superior performance is achieved on widely-used benchmarks including CIFAR-10/100, Omniglot, and ImageNet.

Downloads

Published

2021-05-18

How to Cite

Zhou, F., Li, J., Xie, C., Chen, F., Hong, L., Sun, R., & Li, Z. (2021). MetaAugment: Sample-Aware Data Augmentation Policy Learning. Proceedings of the AAAI Conference on Artificial Intelligence, 35(12), 11097-11105. Retrieved from https://ojs.aaai.org/index.php/AAAI/article/view/17324

Issue

Section

AAAI Technical Track on Machine Learning V