MEA-Defender: A Robust Watermark against Model Extraction Attack

1 code implementation • 26 Jan 2024 • Peizhuo Lv, Hualong Ma, Kai Chen, Jiachen Zhou, Shengzhi Zhang, Ruigang Liang, Shenchen Zhu, Pan Li, Yingjun Zhang

To protect the Intellectual Property (IP) of the original owners over such DNN models, backdoor-based watermarks have been extensively studied.

Model extraction Self-Supervised Learning

DataElixir: Purifying Poisoned Dataset to Mitigate Backdoor Attacks via Diffusion Models

1 code implementation • 18 Dec 2023 • Jiachen Zhou, Peizhuo Lv, Yibing Lan, Guozhu Meng, Kai Chen, Hualong Ma

Dataset sanitization is a widely adopted proactive defense against poisoning-based backdoor attacks, aimed at filtering out and removing poisoned samples from training datasets.

SSL-WM: A Black-Box Watermarking Approach for Encoders Pre-trained by Self-supervised Learning

1 code implementation • 8 Sep 2022 • Peizhuo Lv, Pan Li, Shenchen Zhu, Shengzhi Zhang, Kai Chen, Ruigang Liang, Chang Yue, Fan Xiang, Yuling Cai, Hualong Ma, Yingjun Zhang, Guozhu Meng

Recent years have witnessed tremendous success in Self-Supervised Learning (SSL), which has been widely utilized to facilitate various downstream tasks in Computer Vision (CV) and Natural Language Processing (NLP) domains.

Self-Supervised Learning

DBIA: Data-free Backdoor Injection Attack against Transformer Networks

1 code implementation • 22 Nov 2021 • Peizhuo Lv, Hualong Ma, Jiachen Zhou, Ruigang Liang, Kai Chen, Shengzhi Zhang, Yunfei Yang

In this paper, we propose DBIA, a novel data-free backdoor attack against the CV-oriented transformer networks, leveraging the inherent attention mechanism of transformers to generate triggers and injecting the backdoor using the poisoned surrogate dataset.

Backdoor Attack Image Classification +1