Regularization with Latent Space Virtual Adversarial Training

TL;DR

This paper introduces LVAT, a novel regularization technique that applies adversarial perturbations in the latent space using generative models, leading to improved classification performance in supervised and semi-supervised learning.

Contribution

LVAT extends VAT by injecting perturbations in the latent space with generative models, enabling more effective adversarial sample generation and regularization.

Findings

LVAT outperforms VAT on SVHN and CIFAR-10 datasets.

Latent space perturbations lead to stronger regularization effects.

Using variational auto-encoder and Glow enhances flexibility.

Abstract

Virtual Adversarial Training (VAT) has shown impressive results among recently developed regularization methods called consistency regularization. VAT utilizes adversarial samples, generated by injecting perturbation in the input space, for training and thereby enhances the generalization ability of a classifier. However, such adversarial samples can be generated only within a very small area around the input data point, which limits the adversarial effectiveness of such samples. To address this problem we propose LVAT (Latent space VAT), which injects perturbation in the latent space instead of the input space. LVAT can generate adversarial samples flexibly, resulting in more adverse effects and thus more effective regularization. The latent space is built by a generative model, and in this paper, we examine two different type of models: variational auto-encoder and normalizing flow, specifically Glow. We evaluated the performance of our method in both supervised and semi-supervised learning scenarios for an image classification task using SVHN and CIFAR-10 datasets. In our evaluation, we found that our method outperforms VAT and other state-of-the-art methods.