Biography
Zihao Zhu is currently a PhD student in Data Science at The Chinese University of Hong Kong, Shenzhen (CUHKSZ). His primary research interests are Trustworthy AI, Backdoor Attack and Defense in Deep Learning. He has published related papers in top tier conferences and journals such as NeurIPS, IJCAI, ICASSP and PR.
Interests
- Trustworthy AI
- AI Security & Privacy
- Security in LLMs
Education
-
Ph.D. in Data Science, 2021-
The Chinese University of Hong Kong, Shenzhen
-
MSc in CS, 2018-2021
University of Chinese Academy of Sciences
-
BSc in CS, 2014-2018
China University of Mining and Technology
What’s New
- [2024/01] One paper on LLMs for data cleaning was accepted at ICLR-2024.
- [2024/01] One paper on Few-shot learning was released at arXiv.
- [2024/01] We release Attacks in Adversarial Machine Learning: A Systematic Survey from the Life-cycle Perspective. The draft is available at arXiv.
- [2023/12] We release Defenses in Adversarial Machine Learning: A Survey. The draft is available at arXiv.
- [2023/10] One paper on Robust Backdoor Attack was released. The draft is available at arXiv.
- [2023/07] One paper on Backdoor Sample Selection was released. The draft is available at arXiv.
- [2022/09] One paper on Backdoor Benchmark was accepted at NeurIPS-2022. Code is available on github.
- [2022/01] One paper on VQA was accepted at ICASSP-2022.
- [2021/01] One paper on Machine Reading Comprehension was accepted at ICASSP-2021.
- [2020/12] One paper on VQA was accpeted at Pattern Recognition.
- [2020/04] Two papers on VQA and Visual dialogue were accpeted at IJCAI-2020.
Featured Publications
Recent Publications
Quickly discover relevant content by filtering publications.
(2024).
BackdoorBench: A Comprehensive Benchmark and Analysis of Backdoor Learning.
arXiv 2024.
(2024).
VDC: Versatile Data Cleanser based on Visual-Linguistic Inconsistency by Multimodal Large Language Models.
ICLR 2024.
(2024).
Enhanced Few-Shot Class-Incremental Learning via Ensemble Models.
arXiv 2024.
(2024).
Learning to Optimize Permutation Flow Shop Scheduling via Graph-based Imitation Learning.
AAAI 2024.
(2023).
BlackboxBench: A Comprehensive Benchmark of Black-box Adversarial Attacks.
arXiv 2023.
(2023).
Defenses in Adversarial Machine Learning: A Survey.
arXiv 2023.
(2023).
Robust Backdoor Attack with Visible, Semantic, Sample-Specific, and Compatible Triggers.
arXiv 2023.
(2023).
Boosting Backdoor Attack with A Learnable Poisoning Sample Selection Strategy.
arXiv 2023.
(2023).
Attacks in Adversarial Machine Learning: A Systematic Survey from the Life-cycle Perspective.
arXiv 2023.
(2022).
BackdoorBench: A Comprehensive Benchmark of Backdoor Learning.
NeurIPS 2022.
(2022).
From Shallow to Deep: Compositional Reasoning over Graphs for Visual Question Answering.
ICASSP 2022.
(2021).
MCR-NET: A Multi-Step Co-Interactive Relation Network for Unanswerable Questions on Machine Reading Comprehension.
ICASSP 2021.
(2020).
Cross-Modal Knowledge Reasoning for Knowledge-based Visual Question Answering.
Pattern Recognition.
Experience
Research Assistant
Backdoor Learning, AI security, Adversarial Machine Learning
Research Intern
Large Language Models, Embodied AI
Contact
- 2001 Longxiang Boulevard, Shenzhen, Longgang District 94305
- Teaching Complex B Building, B603