I am a Ph.D. candidate in the Computer Science Department at the University of Wisconsin–Madison, advised by Kassem Fawaz in the Wi-Pi and MadS&P research group. I also work with Nicolas Papernot on adversarial machine learning. Prior to joining UW–Madison, I obtained my Bachelor’s degree in Computer Science from Shanghai University.
My research interest broadly lies in machine learning security and system security. My current works focus on the adversarial robustness of machine learning systems, with the goal of understanding, detecting, and mitigating vulnerabilities in real-world machine learning systems.
I am actively seeking full-time opportunities for Spring and Fall 2024. Here is my CV.
|Apr 20, 2023||Gave a talk about the vulnerabilities of preprocessing in adversarial machine learning at RIKEN-AIP.|
|Oct 11, 2022||Gave a talk about the limitations of stochastic pre-processing defenses (slides).|
|Oct 8, 2022||Recognized as a Top Reviewer (10%) for NeurIPS 2022.|
|Oct 3, 2022||Wrote a blogpost about stochastic pre-processing defenses.|
|Sep 14, 2022||Our paper On the Limitations of Stochastic Pre-processing Defenses was accepted by NeurIPS 2022.|
|May 14, 2022||Our paper The Interplay Between Vulnerabilities in Machine Learning Systems was accepted by ICML 2022 as long presentation (top 2% of all papers).|
|May 2, 2022||Our paper Experimental Security Analysis of the App Model in Business Collaboration Platforms was accepted by USENIX Security 2022.|
arXivSEA: Shareable and Explainable Attribution for Query-based Black-box AttacksPreprint, Aug 2023
NeurIPSOn the Limitations of Stochastic Pre-processing DefensesIn Proceedings of the 36th Conference on Neural Information Processing Systems, Aug 2022
USENIX SecurityExperimental Security Analysis of the App Model in Business Collaboration PlatformsIn 31st USENIX Security Symposium (USENIX Security 22), Aug 2022
ICMLRethinking Image-Scaling Attacks: The Interplay Between Vulnerabilities in Machine Learning SystemsIn Proceedings of the 39th International Conference on Machine Learning, Jul 2022Accepted for Long Presentation (2%)
CVPR WorkshopVariational Autoencoder for Low Bit-rate Image CompressionIn Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR) Workshops, Jul 2018Winner of the 1st Workshop and Challenge on Learned Image Compression.