Learning Loss for Test-Time Augmentation

Abstract

Data augmentation has been actively studied for robust neural networks. Most of the recent data augmentation methods focus on augmenting datasets during the training phase. At the testing phase, simple transformations are still widely used for test-time augmentation. This paper proposes a novel instance-level test-time augmentation that efficiently selects suitable transformations for a test input. Our proposed method involves an auxiliary module to predict the loss of each possible transformation given the input. Then, the transformations having lower predicted losses are applied to the input. The network obtains the results by averaging the prediction results of augmented inputs. Experimental results on several image classification benchmarks show that the proposed instance-aware test-time augmentation improves the model's robustness against various corruptions.

arXiv commentAccepted at NeurIPS 2020

References in corpus (15)

• Very Deep Convolutional Networks for Large-Scale Image Recognition
• mixup: Beyond Empirical Risk Minimization
• Improved Regularization of Convolutional Neural Networks with Cutout
• SGDR: Stochastic Gradient Descent with Warm Restarts
• Aleatoric uncertainty estimation with test-time augmentation for medical image segmentation with convolutional neural networks
• Automatic Classification of Cancerous Tissue in Laserendomicroscopy Images of the Oral Cavity using Deep Learning
• Automatic Brain Tumor Segmentation using Convolutional Neural Networks with Test-Time Augmentation
• Comparing deep neural networks against humans: object recognition when the signal gets weaker
• Examining the Impact of Blur on Recognition by Convolutional Networks
• End-to-End Deep Learning for Steering Autonomous Vehicles Considering Temporal Dependencies
• Unravelling Robustness of Deep Learning based Face Recognition Against Adversarial Attacks
• Augment your batch: better training with larger batches
• Greedy Policy Search: A Simple Baseline for Learnable Test-Time Augmentation
• Quality Resilient Deep Neural Networks
• Propagating Uncertainty in Multi-Stage Bayesian Convolutional Neural Networks with Application to Pulmonary Nodule Detection