Improved Robustness to Open Set Inputs via Tempered Mixup

TL;DR

This paper introduces a simple regularization technique for CNNs that enhances open set robustness, enabling better detection of unseen classes without needing extra background data, and achieves state-of-the-art results.

Contribution

The proposed method is a novel regularization approach that improves open set classification robustness without additional background datasets.

Findings

Achieves state-of-the-art open set classification performance.

Easily scalable to large-scale problems.

Does not require background datasets for training.

Abstract

Supervised classification methods often assume that evaluation data is drawn from the same distribution as training data and that all classes are present for training. However, real-world classifiers must handle inputs that are far from the training distribution including samples from unknown classes. Open set robustness refers to the ability to properly label samples from previously unseen categories as novel and avoid high-confidence, incorrect predictions. Existing approaches have focused on either novel inference methods, unique training architectures, or supplementing the training data with additional background samples. Here, we propose a simple regularization technique easily applied to existing convolutional neural network architectures that improves open set robustness without a background dataset. Our method achieves state-of-the-art results on open set classification baselines and easily scales to large-scale open set classification problems.