Jiachen Li

Ph.D. candidate

jiachen.li[at]rmit.edu.au

I am a Ph.D. candidate at the School of Computing Technologies, RMIT University, advised by Professor Xun Yi, Professor Xiuzhen (Jenny) Zhang, Dr. Xiaoning (Maggie) Liu, Dr. Xiaoyu (Shaw) Xia, and Dr. Bang Wu.

My research explores the security and trustworthiness of Spiking Neural Networks (SNNs), a new generation of energy-efficient brain-inspired AI models. Specifically, I study backdoor attacks, detection, and mitigation techniques under both centralized and federated learning frameworks. My goal is to develop robust, theoretically grounded defenses that require minimal assumptions about the defender’s knowledge, making them more practical for real-world deployment.

Before starting my PhD, I completed a Master of Computer Science and Bachelor of Science at The University of Melbourne.

news

Jul 09, 2025	Our paper Unsupervised Backdoor Detection and Mitigation for Spiking Neural Networks has been accepted by The 28th International Symposium on Research in Attacks, Intrusions and Defenses (RAID 2025).

selected publications

Unsupervised Backdoor Detection and Mitigation for Spiking Neural Networks

Jiachen Li, Bang Wu, Xiaoyu Xia, Xiaoning Liu, Xun Yi, and Xiuzhen Zhang

In Proceedings of the 28th International Symposium on Research in Attacks, Intrusions and Defenses (RAID), 2025

DOI Bib PDF Code

@inproceedings{li2025unsupervised,
  author = {Li, Jiachen and Wu, Bang and Xia, Xiaoyu and Liu, Xiaoning and Yi, Xun and Zhang, Xiuzhen},
  title = {Unsupervised Backdoor Detection and Mitigation for Spiking Neural Networks},
  booktitle = {Proceedings of the 28th International Symposium on Research in Attacks, Intrusions and Defenses (RAID)},
  year = {2025},
  address = {Gold Coast, Australia},
  publisher = {IEEE},
  doi = {10.1109/RAID67961.2025.00025},
  url = {https://arxiv.org/abs/2510.06629},
  eprint = {2510.06629},
  archiveprefix = {arXiv},
  primaryclass = {cs.CR},
  selected = true
}