# Distilling Adversarial Robustness Using Heterogeneous Teachers

Source: [Distilling Adversarial Robustness Using Heterogeneous Teachers](<https://www.sota2.com/research/paper/68fa725a1e209131cea55e13>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2025-10-26T02:24:11.561Z

Authors: Jieren Deng, Aaron Palmer, Rigel Mahmood, Ethan Rathbun, Jinbo Bi, Kaleel Mahmood, Derek Aguiar

Published: 2024-02-23T19:55:13.000Z

## Abstract

Achieving resiliency against adversarial attacks is necessary prior to deploying neural network classifiers in domains where misclassification incurs substantial costs, e.g., self-driving cars or medical imaging. Recent work has demonstrated that robustness can be transferred from an adversarially trained teacher to a student model using knowledge distillation. However, current methods perform distillation using a single adversarial and vanilla teacher and consider homogeneous architectures (i.e., residual networks) that are susceptible to misclassify examples from similar adversarial subspaces. In this work, we develop a defense framework against adversarial attacks by distilling adversarial robustness using heterogeneous teachers (DARHT). In DARHT, the student model explicitly represents teacher logits in a student-teacher feature map and leverages multiple teachers that exhibit low adversarial example transferability (i.e., exhibit high performance on dissimilar adversarial examples). Experiments on classification tasks in both white-box and black-box scenarios demonstrate that DARHT achieves state-of-the-art clean and robust accuracies when compared to competing adversarial training and distillation methods in the CIFAR-10, CIFAR-100, and Tiny ImageNet datasets. Comparisons with homogeneous and heterogeneous teacher sets suggest that leveraging teachers with low adversarial example transferability increases student model robustness.

## Recorded benchmark results

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

## Original sources and implementation

[Original paper](<https://arxiv.org/abs/2402.15586>)

[PDF](<https://arxiv.org/pdf/2402.15586.pdf>)
