I am currently a Ph.D. student in the Department of Computer Science at City University of Hong Kong (CityU), under the supervision of Prof. Xiaohua Jia. Prior to joining CityU, I received my B.E. degree from the Department of Computer Science and Engineering at the Southern University of Science and Technology (SUSTech), supervised by Prof. Xin Yao. My current research interests focus on AI security and safety, as well as VLMs and VLAs in the domain of embodied intelligence.
🔥 News
-
May 2025: 🎉🎉🎉 Happy to announce that our paper CAT: Contrastive Adversarial Training for Evaluating the Robustness of Protective Perturbations in Latent Diffusion Models has been accepted to ICML 2025. Many thanks to my collaborators for their great contributions!
-
Apr. 2025: 🎉🎉🎉 Happy to announce that our paper Improve Fluency Of Neural Machine Translation Using Large Language Models has been accepted to MT Summit 2025. Many thanks to my collaborators for their great contributions!
-
Nov. 2024: 🎉🎉🎉 Happy to announce that our paper Protective Perturbations against Unauthorized Data Usage in Diffusion-based Image Generation has been accepted to CBD 2024. Many thanks to the project leader and my collaborators for their great contributions!
-
Sept. 2024: 🎉🎉🎉 Happy to announce that our paper Intellectual Property Protection of Diffusion Models via the Watermark Diffusion Process has been accepted to WISE 2024. Many thanks to my collaborators for their great contributions!
📝 Selected Publications
Sen Peng, Mingyue Wang, Jianfei He, Jijia Yang, Xiaohua Jia
Project Page | We propose Contrastive Adversarial Training (CAT) to evaluate protective perturbations in LDM customization, revealing their vulnerability by leveraging the latent representation distortion and demonstrating significant effectiveness across extensive experiments.
[WISE 2024] Intellectual Property Protection of Diffusion Models via the Watermark Diffusion Process
Sen Peng, Yufei Chen, Cong Wang, Xiaohua Jia
Project Page | We introduce WDM, a prompt-free watermarking framework for diffusion models that preserves imperceptibility by jointly learning a watermark diffusion process without altering generation quality.
[WWW Journal] Intellectual Property Protection of DNN Models
Sen Peng, Yufei Chen, Jie Xu, Zizhuo Chen, Cong Wang, Xiaohua Jia
We provide a comprehensive survey of emerging techniques for protecting deep neural network intellectual property, including watermarking, fingerprinting, authentication, and inference perturbation, and highlight key challenges and future directions.
📖 Educations
-
Sept. 2020 - Current, Ph.D. in Computer Science, City University of Hong Kong, Hong Kong.
-
Jun. 2019 - Jul. 2019, Summer Workshop at School of Computing, National University of Singapore, Singapore.
-
Sept. 2016 - Jun. 2020, B.E. in Computer Science and Technology, Southern University of Science and Technology, Shenzhen.