Jiaming He

Hi! I'm Jiaming HE (Chinese name: 何嘉铭). I'm currently an undergraduate student majoring in software engineering at Chengdu University of Technology.

I am also an incoming Ph.D student at University of Electronic Science and Technology of China (UESTC), advised by Prof. Hongwei Li.

Email / CV / Google Scholar / Twitter / Github

Taken in Kiyomizu-dera, Kyoto, Japan

Research

My research interest lies in Trustworthy AI. I focus on building robust and reliable methods to implement deep learning system in physical world.

Long term goal: Building Trustworthy AGI.

Misc

I am a loyal fan of League of Legends🎮️, and Football. I support the best football team in London - Chelsea. Let's keep blue flag flying high!💙 Do not hesitate to contact me if you are interested in collaboration or communication (any topics: Research, Football, Game, Hold'em, etc.). Let's make something great!

News

[3/2025] Our paper on Text-to-image Generation Backdoor is accepted to ICME 2025!
[1/2025] Our paper accepted to AAAI 2025 has been seleced for Oral Presentation (Top 4%)!
[12/2024] Our paper on defending privacy in RAG systems is accepted to ICASSP 2025!
[12/2024] Our paper investigate in condition backdoor attack on LLMs is accepted to AAAI 2025!
[09/2024] I will be joining UESTC as a first-year Ph.D student in 2025!
[01/2024] Our paper on Image-scaling attacks is accepted to ICC 2024!

Selected Publications

	Watch Out for Your Guidance on Generation! Exploring Conditional Backdoor Attacks against Large Language Models. Jiaming He, Wenbo Jiang, Guanyu Hou, Wenshu Fan, Rui Zhang, Hongwei Li AAAI Conference on Artificial Intelligence (AAAI) (CCF-A) (Oral Presentation), 2025 Investigate the conditional backdoor attack against LLMs by exploiting generation conditions during inference.
	Weaponizing Tokens: Backdooring Text-to-Image Generation via Token Remapping. Jiaming He, Wenbo Jiang, Guanyu Hou, Qiyang Song, Ji Guo, Hongwei Li, IEEE International Conference on Multimedia & Expo (ICME) (CCF-B), 2025 Investigate a trainin-free backdoor attack on Text-to-image generation.
	PRESS: Defending Privacy in Retrieval-augmented Generation via Embedding Space Shifting. Jiaming He, Cheng Liu, Guanyu Hou, Wenbo Jiang, Jiacheng Li IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP) (CCF-B)*, 2025 Applying purposive embedding space shifting for defending privacy in RAG-based LLMs.
	MTISA: Multi-Target Image-Scaling Attack. Jiaming He, Hongwei Li, Wenbo Jiang, Yuan Zhang IEEE International Conference on Communications (ICC) (CCF-C) (Oral Presentation), 2024 Introduce multiple targets in image-scaling attacks.
	Data Stealing Attacks against Large Language Models via Backdooring Jiaming He, Guanyu Hou, Xinyue Jia, Yangyang Chen, Wenqi Liao, Yinhang Zhou, Rang Zhou Electronics (JCR-Q2) , 2024 Propose a data stealing attacks against LLMs via backdoor injection
	Combinational Backdoor Attack against Customized Text-to-Image Models Wenbo Jiang, Jiaming He, Hongwei Li, Guowen Xu, Rui Zhang, Hanxiao Chen, Meng Hao, Haomiao Yang, arXiv , 2024 (In Submmition) Explore combinational backdoor attack on customized Text-to-image generative models.
	Backdoor Attacks against Image-to-Image Networks Wenbo Jiang, Hongwei Li, Jiaming He, Rui Zhang, Guowen Xu, Tianwei Zhang, Rongxing Lu arXiv , 2023 (In Submmition) Backdoor Attacks against Image-to-Image Networks

Academic Services

Reviewer:

Annual Conference on Neural Information Processing Systems (NIPS) 2025
IEEE International Conference on Multimedia & Expo (ICME) 2025
IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP) 2025
IEEE International Workshop on Machine Learning for Signal Processing (MLSP) 2025
International Joint Conference on Neural Networks (IJCNN) 2025