Occlusions for Effective Data Augmentation in Image Classification

TL;DR

This paper demonstrates that using occlusions as a data augmentation technique, combined with batch augmentation, improves image classification performance on large-scale datasets like ImageNet, especially for high-capacity models.

Contribution

It introduces a simple occlusion-based data augmentation method that enhances ImageNet classification results and provides a way to assess model robustness.

Findings

Occlusions improve classification accuracy on ImageNet for ResNet50.

Varying occlusion levels helps analyze neural network robustness.

Batch augmentation enhances the effectiveness of occlusion-based data augmentation.

Abstract

Deep networks for visual recognition are known to leverage "easy to recognise" portions of objects such as faces and distinctive texture patterns. The lack of a holistic understanding of objects may increase fragility and overfitting. In recent years, several papers have proposed to address this issue by means of occlusions as a form of data augmentation. However, successes have been limited to tasks such as weak localization and model interpretation, but no benefit was demonstrated on image classification on large-scale datasets. In this paper, we show that, by using a simple technique based on batch augmentation, occlusions as data augmentation can result in better performance on ImageNet for high-capacity models (e.g., ResNet50). We also show that varying amounts of occlusions used during training can be used to study the robustness of different neural network architectures.