This paper investigates a new challenging problem called
defensive few-shot learningin order to learn a robust few-shot model against adversarial attacks. Simply applying the existing adversarial defense methods to few-shot learning cannot effectively solve this problem. This is because the commonly assumed sample-level distribution consistency between the training and test sets can no longer be met in the few-shot setting. To address this situation, we develop a general defensive few-shot learning (DFSL) frameworkto answer the following two key questions: (1) how to transfer adversarial defense knowledge from one sample distribution to another? (2) how to narrow the distribution gap between clean and adversarial examples under the few-shot setting? To answer the first question, we propose an episode-based adversarial training mechanismby assuming a task-level distribution consistencyto better transfer the adversarial defense knowledge. As for the second question, within each few-shot task, we design two kinds of distribution consistency criteria to narrow the distribution gap between clean and adversarial examples from the feature-wiseand prediction-wiseperspectives, respectively. Extensive experiments demonstrate that the proposed framework can effectively make the existing few-shot models robust against adversarial attacks. Code is available at https://github.com/WenbinLee/DefensiveFSL.git.