Litcius/Paper detail

MetaAugment: Sample-Aware Data Augmentation Policy Learning

Fengwei Zhou, Jiawei Li, Chuanlong Xie, Fei Chen, Lanqing Hong, Rui Sun, Zhenguo Li

2021Proceedings of the AAAI Conference on Artificial Intelligence19 citationsDOIOpen Access PDF

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.

Topics & Concepts

Task (project management)Computer scienceConvergence (economics)Sample (material)Image (mathematics)Transformation (genetics)Set (abstract data type)Artificial intelligenceMachine learningRate of convergencePolicy learningTraining setKey (lock)EngineeringBiochemistryEconomic growthProgramming languageChromatographyGeneComputer securityEconomicsChemistrySystems engineeringAdvanced Neural Network ApplicationsDomain Adaptation and Few-Shot LearningMedical Image Segmentation Techniques