NT-ML: Backdoor Defense via Non-target Label Training and Mutual Learning

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Huo, Wenjie, Wolter, Katinka
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866912525872791552
author Huo, Wenjie
Wolter, Katinka
author_facet Huo, Wenjie
Wolter, Katinka
contents Recent studies have shown that deep neural networks (DNNs) are vulnerable to backdoor attacks, where a designed trigger is injected into the dataset, causing erroneous predictions when activated. In this paper, we propose a novel defense mechanism, Non-target label Training and Mutual Learning (NT-ML), which can successfully restore the poisoned model under advanced backdoor attacks. NT aims to reduce the harm of poisoned data by retraining the model with the outputs of the standard training. At this stage, a teacher model with high accuracy on clean data and a student model with higher confidence in correct prediction on poisoned data are obtained. Then, the teacher and student can learn the strengths from each other through ML to obtain a purified student model. Extensive experiments show that NT-ML can effectively defend against 6 backdoor attacks with a small number of clean samples, and outperforms 5 state-of-the-art backdoor defenses.
format Preprint
id arxiv_https___arxiv_org_abs_2508_05404
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle NT-ML: Backdoor Defense via Non-target Label Training and Mutual Learning
Huo, Wenjie
Wolter, Katinka
Machine Learning
Recent studies have shown that deep neural networks (DNNs) are vulnerable to backdoor attacks, where a designed trigger is injected into the dataset, causing erroneous predictions when activated. In this paper, we propose a novel defense mechanism, Non-target label Training and Mutual Learning (NT-ML), which can successfully restore the poisoned model under advanced backdoor attacks. NT aims to reduce the harm of poisoned data by retraining the model with the outputs of the standard training. At this stage, a teacher model with high accuracy on clean data and a student model with higher confidence in correct prediction on poisoned data are obtained. Then, the teacher and student can learn the strengths from each other through ML to obtain a purified student model. Extensive experiments show that NT-ML can effectively defend against 6 backdoor attacks with a small number of clean samples, and outperforms 5 state-of-the-art backdoor defenses.
title NT-ML: Backdoor Defense via Non-target Label Training and Mutual Learning
topic Machine Learning
url https://arxiv.org/abs/2508.05404