I am an Associate Professor at Xi’an Jiaotong University (XJTU), China. I was a postdoc at CISPA Helmholtz Center for Information Security, Germany, hosted by Prof. Michael Backes.
I received my PhD from Radboud University, Netherlands, supervised by Prof. Martha Larson, with my PhD thesis entitled “Rethinking Realism: Towards More Transferable and Less Suspicious Adversarial Images”.
Research Interests
- Security of Machine Learning: Foundations of attacks/defenses
(ICML'24, TIFS'24, ICML'23, ICLR'23, TIFS'23, CCS'22, NeurIPS'21, CVPR'20, BMVC'20, ICMR'19)
- Security of Machine Learning: Applications of attacks/defenses in large (multimodal) models, autonomous driving...
(CVPR'24, TIFS'24, TOPS'24, NAACL Findings'24, UMAP'21)
- Machine Learning for Security: Detecting AI-generated multimedia, preventing inference of sensitive information...
(NeurIPS'24, TIFS'24, NDSS'21, MM'18)
Resources
Services
- Poster Session Co-Chair of MM 2019
- Area Chair of NeurIPS, Senior Program Committee of AAAI
- Program Committee of ICLR, ICML, NeurIPS, CVPR, ICCV, ECCV, ACL, AAAI, IJCAI, AISTATS, BMVC, FAccT
- Journal Reviewer of TPAMI, TIFS, TKDE, TDSC, TCSS, IJCV, PR
- Task Co-Organizer of Pixel Privacy and Multimedia Satellite at MediaEval Workshop 2018-2020
Honors & Awards
Invited Talks
- Security and Privacy Risks of AI Large Models, ICIG2023, 2023-09-23.
- Adversarial Examples and Data Poisons, Saarland University (Guest Lecture), 2023-06-19.
- Computer Vision in Adversarial Scenarios, LIS - Ecole Centrale Marseille, 2023-03-03. [Slides]
- Computer Vision against Adversarial Perturbations, Hong Kong PolyU, 2022-12-16. [Slides]
- Transferable and Stealthy Adversarial Images, Alibaba Turing Lab, 2022-03-03. [Slides]
- Transferability of Targeted Attacks, AI TIME, 2022-02-17. [Video (in Chinese)] [Slides]
Miscellaneous
- I like simple yet effective (research) ideas.
- I love music, particularly Chinese music.
- I respect Dr. Nicholas Carlini for his long-term dedication to rightful evaluations.