Zihan Wang

About Me

I began my Ph.D. studies in 2023 at the School of Electrical Engineering and Computer Science, the University of Queensland, under the supervision of A/Prof. Guangdong Bai and Dr. Jason Xue. My research focuses on tackling real-world security & privacy issues of machine learning systems in a formally verifiable manner. My works have been published in leading conferences and journals, including IEEE S&P, ACM CCS, USENIX Security, NeurIPS, WACV, and IEEE TCSS. I am also the HDR representative of the discipline of Cyber Security and Software Engineering (CSS).

Research Interests

Trustworthy and Responsible AI
- Neural Network Usage Control: CoreLocker@IEEE S&P24
- Purpose Limitation: AlgoSpec@USENIX Security24
- Neural Network Verification
Privacy-preserving Machine Learning
- Gradient Inversion Attack: Grab@CCS24
- Membership Inference Attack: M⁴I@NeurIPS22
- Data Hiding: IEEE TCSS23
Efficient Machine Learning:
- Model Pruning: CoreLocker@IEEE S&P24
- Knowledge Distillation: BPKD@WACV24

News

[Aug. 2024] Our paper on GIA in language model is accepted by CCS’24 (Core A*, Big Four).
[Mar. 2024] Our paper on NN usage control is accepted by IEEE S&P’24 (Core A*, Big Four).
[Feb. 2024] Our paper on purpose limitation is accepted by Security’24 (Core A*, Big Four).
[Dec. 2023] Our paper on deep data hiding is accepted by IEEE TCSS (JCR Q1).
[Oct. 2023] Our paper on knowledge distillation is accepted by WACV’24 (Core A).
[Aug. 2023] Our paper on formalizing neural network perturbation is accepted by ICFEM’23.
[Dec. 2022] I obtained my Bachelor of Comp. Sci. (Advanced) from the University of Adelaide!
[Dec. 2022] Excited to have presented at my first conference @ NeurIPS in New Orleans! [Photo]
[Sep. 2022] Our paper on multi-modal model MIA is accepted by NeurIPS’22 (Core A*).

Invited Talks

[May. 2024] Neuron-level Usage Control for AI Models, School of Computing, National University of Singapore (NUS).

Selected Publications

S&P'24

CoreLocker: Neuron-level Usage Control

Zihan Wang, Zhongkui Ma, Xinguo Feng, Ruoxi Sun, Hu Wang, Minhui Xue, Guangdong Bai.

IEEE Symposium on Security and Privacy (S&P), Mar. 2024.

PDF Website

BibTeX

@inproceedings{wang2024corelocker,
  title={CoreLocker: Neuron-level Usage Control},
  author={Wang, Zihan and Ma, Zhongkui and Feng, Xinguo and Sun, Ruoxi and Wang, Hu and Xue, Minhui and Bai, Guangdong.},
  booktitle={IEEE Symposium on Security and Privacy (S\&P)},
  doi = {10.1109/SP54263.2024.00182},
  url = {https://doi.ieeecomputersociety.org/10.1109/SP54263.2024.00182},
  pages = {222-222},
  year={2024}
}

USENIX'24

Being Transparent is Merely the Beginning: Enforcing Purpose Limitation with Polynomial Approximation

Shuofeng Liu, Zihan Wang, Minhui Xue, Long Wang, Yuanchao Zhang, Guangdong Bai.

USENIX Security, Feb. 2024.

PDF

BibTeX

@inproceedings{liu2024purpose,
  title={Being Transparent is Merely the Beginning: Enforcing Purpose Limitation with Polynomial Approximation},
  author={Liu, Shuofeng and Wang, Zihan and Xue, Minhui and Wang, Long and Zhang, Yuanchao and Bai, Guangdong.},
  journal={USENIX Security},
  year={2024}
}

CCS'24
Uncovering Gradient Inversion Risks in Practical Language Model Training

Xinguo Feng, Zhongkui Ma, Zihan Wang, Chegne Eu Joe, Mengyao Ma, Alsharif Abuadbba, Guangdong Bai.

ACM Computer and Communications Security Conference (CCS), Aug. 2024.
PDF
BibTeX

@inproceedings{ }

TCSS

Data Hiding With Deep Learning: A Survey Unifying Digital Watermarking and Steganography

Zihan Wang, Olivia Byrnes, Hu Wang, Ruoxi Sun, Congbo Ma, Huaming Chen, Qi Wu, Minhui Xue.

IEEE Transactions on Computational Social Systems (TCSS), Dec. 2023.

PDF

BibTeX

@article{wang2023data,
  title={Data hiding with deep learning: a survey unifying digital watermarking and steganography},
  author={Wang, Zihan and Byrnes, Olivia and Wang, Hu and Sun, Ruoxi and Ma, Congbo and Chen, Huaming and Wu, Qi and Xue, Minhui},
  journal={IEEE Transactions on Computational Social Systems},
  year={2023},
  publisher={IEEE}
}

NeurIPS'22

M⁴I: Multi-modal Models Membership Inference

Pingyi Hu, Zihan Wang*, Ruoxi Sun, Hu Wang, Minhui Xue.

Neural Information Processing Systems (NeurIPS), Sep. 2022.

PDF

BibTeX

@inproceedings{hu2022m,
  title={M $\^{} 4$ I: Multi-modal Models Membership Inference},
  author={Hu, Pingyi and Wang, Zihan and Sun, Ruoxi and Wang, Hu and Xue, Minhui},
  journal={Advances in Neural Information Processing Systems},
  volume={35},
  pages={1867--1882},
  year={2022}
}

WACV'24

BPKD: Boundary Privileged Knowledge Distillation For Semantic Segmentation

Liyang Liu, Zihan Wang, Minh Hieu Phan, Bowen Zhang, Jinchao Ge, Yifan Liu.

IEEE/CVF Winter Conference on Applications of Computer Vision (WACV), Oct. 2023.

PDF

BibTeX

@inproceedings{liu2024bpkd,
  title={BPKD: Boundary Privileged Knowledge Distillation For Semantic Segmentation},
  author={Liu, Liyang and Wang, Zihan and Phan, Minh Hieu and Zhang, Bowen and Ge, Jinchao and Liu, Yifan},
  booktitle={Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision},
  pages={1062--1072},
  year={2024}
}