AutoAugment: Learning Augmentation Policies from Data

24 May 2018Ekin D. CubukBarret ZophDandelion ManeVijay VasudevanQuoc V. Le

Data augmentation is an effective technique for improving the accuracy of modern image classifiers. However, current data augmentation implementations are manually designed... (read more)

PDF Abstract

Results from the Paper


TASK DATASET MODEL METRIC NAME METRIC VALUE GLOBAL RANK RESULT LEADERBOARD
Fine-Grained Image Classification Caltech-101 AutoAugment Top-1 Error Rate 13.07% # 2
Image Classification CIFAR-100 PyramidNet+ShakeDrop Percentage correct 89.3 # 6
Fine-Grained Image Classification FGVC Aircraft AutoAugment Top-1 Error Rate 7.33 # 1
Accuracy 92.67% # 14
Fine-Grained Image Classification Oxford 102 Flowers AutoAugment Top-1 Error Rate 4.64% # 4
Accuracy 95.36% # 10
Fine-Grained Image Classification Oxford-IIIT Pets AutoAugment Top-1 Error Rate 11.02% # 8
Accuracy 88.98% # 9
Fine-Grained Image Classification Stanford Cars AutoAugment Top-1 Error Rate 5.2 # 1
Accuracy 94.8% # 8
Image Classification SVHN AutoAugment Percentage error 1.02 # 1

Methods used in the Paper