Greedy Policy Search: A Simple Baseline for Learnable Test-Time Augmentation

Abstract

Test-time data augmentation$-$averaging the predictions of a machine learning model across multiple augmented samples of data$-$is a widely used technique that improves the predictive performance. While many advanced learnable data augmentation techniques have emerged in recent years, they are focused on the training phase. Such techniques are not necessarily optimal for test-time augmentation and can be outperformed by a policy consisting of simple crops and flips. The primary goal of this paper is to demonstrate that test-time augmentation policies can be successfully learned too. We introduce greedy policy search (GPS), a simple but high-performing method for learning a policy of test-time augmentation. We demonstrate that augmentation policies learned with GPS achieve superior predictive performance on image classification problems, provide better in-domain uncertainty estimation, and improve the robustness to domain shift.

References in corpus (13)

• Very Deep Convolutional Networks for Large-Scale Image Recognition
• mixup: Beyond Empirical Risk Minimization
• Wide Residual Networks
• On Calibration of Modern Neural Networks
• Weight Uncertainty in Neural Networks
• Can You Trust Your Model's Uncertainty? Evaluating Predictive Uncertainty Under Dataset Shift
• Aleatoric uncertainty estimation with test-time augmentation for medical image segmentation with convolutional neural networks
• AugMix: A Simple Data Processing Method to Improve Robustness and Uncertainty
• Population Based Augmentation: Efficient Learning of Augmentation Policy Schedules
• Benchmarking Neural Network Robustness to Common Corruptions and Surface Variations
• Pitfalls of In-Domain Uncertainty Estimation and Ensembling in Deep Learning
• Adversarial Examples Improve Image Recognition
• Mixup Inference: Better Exploiting Mixup to Defend Adversarial Attacks

Cited by in corpus (6)

• A Review of Uncertainty Quantification in Deep Learning: Techniques, Applications and Challenges
• MEMO: Test Time Robustness via Adaptation and Augmentation
• Look Before You Leap: An Exploratory Study of Uncertainty Measurement for Large Language Models
• Learning Loss for Test-Time Augmentation
• Mean Embeddings with Test-Time Data Augmentation for Ensembling of Representations
• Using Artificial Populations to Study Psychological Phenomena in Neural Models