SSL-OTA: Unveiling Backdoor Threats in Self-Supervised Learning for Object Detection

no code implementations • 30 Dec 2023 • Qiannan Wang, Changchun Yin, Liming Fang, Lu Zhou, Zhe Liu, Run Wang, Chenhao Lin

The extensive adoption of Self-supervised learning (SSL) has led to an increased security threat from backdoor attacks.

Backdoor Attack Data Poisoning +5

GhostEncoder: Stealthy Backdoor Attacks with Dynamic Triggers to Pre-trained Encoders in Self-supervised Learning

no code implementations • 1 Oct 2023 • Qiannan Wang, Changchun Yin, Zhe Liu, Liming Fang, Run Wang, Chenhao Lin

Pre-trained image encoders can serve as feature extractors, facilitating the construction of downstream classifiers for various tasks.

Backdoor Attack Image Steganography +1

SSL-Auth: An Authentication Framework by Fragile Watermarking for Pre-trained Encoders in Self-supervised Learning

no code implementations • 9 Aug 2023 • Xiaobei Li, Changchun Yin, Liyue Zhu, Xiaogang Xu, Liming Fang, Run Wang, Chenhao Lin

Self-supervised learning (SSL), a paradigm harnessing unlabeled datasets to train robust encoders, has recently witnessed substantial success.

Self-Supervised Learning

PromptAttack: Prompt-based Attack for Language Models via Gradient Search

no code implementations • 5 Sep 2022 • Yundi Shi, Piji Li, Changchun Yin, Zhaoyang Han, Lu Zhou, Zhe Liu

Therefore, in this paper, we propose a malicious prompt template construction method (\textbf{PromptAttack}) to probe the security performance of PLMs.