About Me

Hello, I am Peizhuo Lv (吕培卓 in Chinese), a Research Fellow at Nanyang Technological University, working with Prof. Xiaofeng Wang, Assoc. Prof. Tianwei Zhang, and Asst. Prof. Wei Dong. I received my Ph. D. degree from Institute of Information Engineering, Chinese Academy of Sciences, where I was supervised by Prof. Kai Chen and Assoc. Prof. Shengzhi Zhang from 2020 to 2025. Before that, I received my B.S. degree from the School of Computer Science and Technology, Xidian University from 2016 to 2020.

My main research interests include AI security, Neural Network watermark, Backdoor attack, Data privacy, and Adversarial attack. Now, I am interested in the security problems of large models.

Emails: lvpeizhuo at gmail.com; peizhuo.lyu at ntu.edu.sg

Research Areas:

Backdoor Attacks and Defense: USENIX Security'23, USENIX Security'23, AAAI'24, ICME'23, ECAI'23
Deep Neural Network Watermarks: S&P'24, NDSS'24, TDSC'23, 计算机研究与发展'21
Large Model Security and Privacy: RAID'24, AAAI'25, ICASSP'25

News:

01/2025: I received my Ph.D. degree from UCAS. I am grateful to Prof. Kai Chen and Assoc. Prof. Shengzhi Zhang for their supervision.
12/2024: One paper "A MODEL STEALING ATTACK AGAINST MULTI-EXIT NETWORKS" is accepted by the 2025 IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP 2025). Congratulations to Pan.
12/2024: One paper "RepeatLeakage: leak prompts from repeating as Large language model is a good repeater" is accepted by Thirty-Ninth AAAI Conference on Artificial Intelligence (AAAI-25). Congratulations to Yu, and Lijie.
11/2024-now: I serve as a reviewer for IEEE Transactions on Information Forensics and Security (IEEE TIFS).
07/2024: One paper "KGDist: A Prompt-Based Distillation Attack against LMs Augmented with Knowledge Graphs" is accepted by the 27th International Symposium on Research in Attacks, Intrusions and Defenses (RAID 2024). Congratulations to Hualong, Jiachen.
01/2024: I am invited by SHUZIHUANYU to give a talk about deep learning watermarks. The video of the talk is available.
12/2023: One paper "DataElixir: Purifying Poisoned Dataset to Mitigate Backdoor Attacks via Diffusion Models" is accepted by Thirty-Eighth AAAI Conference on Artificial Intelligence (AAAI 2024). Congratulations to Jiachen, Yibing, and Hualong.
11/2023: One paper "SSL-WM: A Black-Box Watermarking Approach for Encoders Pre-trained by Self-Supervised Learning" is accepted by The Network and Distributed System Security Symposium (NDSS) 2024.
10/2023: One paper "MEA-Defender: A Robust Watermark against Model Extraction Attack" is accepted by IEEE Symposium on Security and Privacy (S&P) 2024.
10/2023: I serve as a reviewer for AAAI 2024.
07/2023: One paper "Invisible Backdoor Attacks Using Data Poisoning in Frequency Domain" is accepted by ECAI 2023. Congratulations to Chang.
06/2023: I serve as a reviewer for ECAI 2023.

05/2023: One paper "Aliasing Backdoor Attacks on Pre-trained Models" is accepted by USENIX Security 2023. Congratulations to Chengan.
01/2023: One paper "A Data-free Backdoor Injection Approach in Neural Networks" is accepted by USENIX Security 2023.
01/2023: One paper "A Robustness-Assured White-Box Watermark in Neural Networks" is accepted as a regular paper by IEEE TDSC.
07/2022-now: I serve as a reviewer for IEEE Transactions on Dependable and Secure Computing (IEEE TDSC).

Honors & Awards:

National Scholarship for Graduate Students, Ministry of Education of the People’s Republic of China, 2023.
Director Scholarship (Excellent Prize), Institute of Information Engineering, Chinese Academy of Sciences, 2023.
Excellent Student Award, University of Chinese Academy of Sciences, 2023
“Call to Arms” award, 26th European Conference on Artificial Intelligence (ECAI), 2023

Profile Links:

Google Scholar: https://scholar.google.com/citations?user=i6Re3L0AAAAJ&hl=en&oi=ao
Semantic Scholar: https://www.semanticscholar.org/author/Peizhuo-Lv/117426360

Moreover, I am always willing to project co-operations. Feel free to drop me an email if you have any ideas or suitable opportunities to discuss!

Please click the headers above for more information about me or contact me, thanks.