When and Why Test-Time Augmentation Works

11/23/2020
by   Divya Shanmugam, et al.
6

Test-time augmentation (TTA)—the aggregation of predictions across transformed versions of a test input—is a common practice in image classification. In this paper, we present theoretical and experimental analyses that shed light on 1) when test time augmentation is likely to be helpful and 2) when to use various test-time augmentation policies. A key finding is that even when TTA produces a net improvement in accuracy, it can change many correct predictions into incorrect predictions. We delve into when and why test-time augmentation changes a prediction from being correct to incorrect and vice versa. Our analysis suggests that the nature and amount of training data, the model architecture, and the augmentation policy all matter. Building on these insights, we present a learning-based method for aggregating test-time augmentations. Experiments across a diverse set of models, datasets, and augmentations show that our method delivers consistent improvements over existing approaches.

READ FULL TEXT

page 5

page 6

page 8

06/27/2022

Improved Text Classification via Test-Time Augmentation

Test-time augmentation – the aggregation of predictions across transform...
05/13/2021

Adaptive Test-Time Augmentation for Low-Power CPU

Convolutional Neural Networks (ConvNets) are trained offline using the f...
07/19/2018

Test-time augmentation with uncertainty estimation for deep learning-based medical image segmentation

Data augmentation has been widely used for training deep learning system...
02/21/2020

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

Test-time data augmentation—averaging the predictions of a machine learn...
10/29/2021

Boosting Anomaly Detection Using Unsupervised Diverse Test-Time Augmentation

Anomaly detection is a well-known task that involves the identification ...
07/11/2022

MT-Net Submission to the Waymo 3D Detection Leaderboard

In this technical report, we introduce our submission to the Waymo 3D De...
06/22/2021

Test-time Collective Prediction

An increasingly common setting in machine learning involves multiple par...