About
Zhao Zhe (George) is a Research Engineer at RealAI, whose research interest lies in the area of trustworthy AI, software testing and verification, such as the robustness of large language models (LLMs), adversarial attack and defense, neural network testing, etc. He received the PhD degree from System and Software Security Lab (S3L) in School of Information Science and Technology, ShanghaiTech university, under supervision of Prof. Song Fu.
Contact
中文简历(更新日期:2022.08), Google Scholar, Github
wechat=Persist_GeorgeZ
name="zhaozhe"
suffix="1"
email=${name}${suffix}@shanghaitech.edu.cn
Latest News
- Ph.D. thesis defense: check!
- Our paper on query-free black-box adversarial attack against speaker recognition systems is accepted by USENIX 2023.
- Win 2nd Place Prize in MLSec Face Recognition Challenge.
- Our paper on using adversarial attacks to accelerate neural network verification is accepted by SAS 2022.
- Our paper "QVIP: An ILP-based Formal Verification Approach for Quantized Neural Networks" is accepted by ASE 2022.
- I am currently a research assistant / visiting student in Singapore Management University, RISE Lab, co-advised by Prof. Sun Jun
- Serve on the Artifact Evaluation Committee (AEC) of OSDI, Usenix ATC and ISSTA 2022.
- Award China national scholarship.
Publications
QFA2SR: Query-Free Adversarial Transfer Attacks to Speaker Recognition Systems
Guangke Chen, Yedi Zhang, Zhe Zhao, Fu Song
USENIX, CCF-A
CLEVEREST: Accelerating CEGAR-based Neural Network Verification via Adversarial Attacks
Zhe Zhao, Yedi Zhang, Guangke Chen, Fu Song, Taolue Chen and Jiaxiang Liu
SAS
2022, CCF-B
Precise Quantitative Analysis of Binarized Neural Networks: A BDD-based Approach
Yedi Zhang, Zhe Zhao, Guangke Chen, Fu Song, Min Zhang, Taolue Chen
TOSEM, CCF-A
QVIP: An ILP-based Formal Verification Approach for Quantized Neural Networks
Yedi Zhang, Zhe Zhao, Guangke Chen, Fu Song, Min Zhang, Taolue Chen, Jun Sun
ASE
2022, CCF-A
AS2T: Arbitrary Source-To-Target Adversarial Attack on Speaker Recognition Systems
Guangke Chen, Zhe Zhao, Fu Song, Sen Chen, Lingling Fan, Yang Liu
TDSC, CCF-A
Attack as Defense: Characterizing Adversarial Examples using Robustness
Zhe Zhao, Guangke Chen, Jingyi Wang, Yiwei Yang, Fu Song, Jun Sun
ISSTA
2021, CCF-A
[code] [slides]
BDD4BNN: A BDD-based Quantitative Analysis Framework for Binarized Neural Networks
Yedi Zhang, Zhe Zhao, Guangke Chen, Fu Song, Taolue Chen
CAV
2021, CCF-A
Who is Real Bob? Adversarial Attacks on Speaker Recognition Systems
Guangke Chen, Sen Chen, Lingling Fan, Xiaoning Du, Zhe Zhao, Fu Song, Yang Liu
IEEE S&P, Oakland
2021, CCF-A
[code] [slides] [video] [website] [blog]
Taking Care of The Discretization Problem: A Comprehensive Study of the Discretization Problem and A Black-Box Adversarial Attack in Discrete Integer Domain
Lei Bu*, Zhe Zhao*, Yuchao Duan, Fu Song
TDSC, CCF-A
(Co-first author*)
[code]
Award & Scholarship
2nd Place Prize in MLSec Face Recognition Challenge, 📃 write-up 2022
5th Place in CVPR 2022 AISafety Challenge Track2 (AEs detection competition) 2022
Merit Prize in OPPO 2021 Security AI Challenge2021
China National Scholarship 2021
3rd Place in ACM MM 2021 Robust Logo Detection Competition 2021
3rd Place in CVPR 2021 White-box Adversarial Attacks Competition, 📃 paper 2021
1st Place of Baidu PaddlePaddle AI Adversarial Attack Contest, 📃 write-up in Chinese2019
SIST Excellent Teaching Assistant 2019
Excellent Graduate of Ocean University of China 2016
China National Encouragement Scholarship 2015
Teaching
Teaching Assistant for Software Engineering 2019
Service
Committee: OSDI and Usenix ATC AEC (2022), ISSTA AEC (2022)
Reviewer: ISSRE (2022, 2021), ICECCS (2022, 2020, 2019), ICICS (2022, 2021), CAV (2020), AsianHOST (2022)
Student Volunteer: ISSTA (2019)