Adversarial Active Learning for Deep Networks: a Margin Based Approach

TL;DR

This paper introduces an adversarial active learning method for deep neural networks that uses adversarial examples to approximate decision boundary proximity, leading to faster training convergence.

Contribution

It presents a novel margin-based active learning strategy leveraging adversarial examples to improve data efficiency in training deep networks.

Findings

Faster convergence of CNNs on multiple datasets.

Adversarial active queries reduce the number of annotations needed.

Method outperforms traditional uncertainty-based active learning.

Abstract

We propose a new active learning strategy designed for deep neural networks. The goal is to minimize the number of data annotation queried from an oracle during training. Previous active learning strategies scalable for deep networks were mostly based on uncertain sample selection. In this work, we focus on examples lying close to the decision boundary. Based on theoretical works on margin theory for active learning, we know that such examples may help to considerably decrease the number of annotations. While measuring the exact distance to the decision boundaries is intractable, we propose to rely on adversarial examples. We do not consider anymore them as a threat instead we exploit the information they provide on the distribution of the input space in order to approximate the distance to decision boundaries. We demonstrate empirically that adversarial active queries yield faster convergence of CNNs trained on MNIST, the Shoe-Bag and the Quick-Draw datasets.