Jiahuan Long

Assistant Professor (Hundred Talents Program), Ph.D. Supervisor

Email:: jiahuanlong(at)szu.edu.cn
Affiliation:: School of Artificial Intelligence, Shenzhen University
Office:: Room 822, Huiyan Building, Yuehai Campus
Research:: Adversarial machine learning, Embodied AI security, VLLM security

I am an Assistant Professor (Hundred Talents Program) at Shenzhen University. Before joining Shenzhen University, I completed my Ph.D. from Shanghai Jiao Tong University in 2026, supervised by Prof. Xiaoqian Chen and Prof. Chao Ma. During 2022 to 2026, I was a joint-training PhD student at the Academy of Military Sciences, supervised by Prof. Wen Yao and Prof. Tingsong Jiang. Before that, I received my master's degrees from the University of Electronic Science and Technology of China (UESTC) and the University of Edinburgh in 2020 and 2021, respectively. At UESTC, I was supervised by Prof. Ke Zhang and Prof. Xiaofen Wang. My research interests include adversarial machine learning, embodied AI security, and multimodal large language model security.

I am recruiting Ph.D. and master's students to work on AI security for Fall 2027. Undergraduate students interested in gaining research experience are also welcome to join our group. Feel free to contact me!

News

2026.08 Excited to join the School of Artificial Intelligence at Shenzhen University! 🎉

Selected Papers

Thermally Activated Dual-Modal Adversarial Clothing against AI Surveillance Systems

Jiahuan Long, Tingsong Jiang, Hanqing Liu, Chao Ma, Weien Zhou, Yang Yang, and Wen Yao

CVPR 2026 (Highlight, CCF A)

arXiv Video 1 Video 2 Video 3

@inproceedings{long2026thermalclothing,
  title={Thermally Activated Dual-Modal Adversarial Clothing against AI Surveillance Systems},
  author={Long, Jiahuan and Jiang, Tingsong and Liu, Hanqing and Ma, Chao and Zhou, Weien and Yang, Yang and Yao, Wen},
  booktitle={Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)},
  year={2026}
}

Overview of thermally activated dual-modal adversarial clothing

CDUPatch: Color-Driven Universal Adversarial Patch Attack for Dual-Modal Visible-Infrared Detectors

Jiahuan Long, Wen Yao, Tingsong Jiang, Jiacheng Hou, Shuai Jia, Junqi Wu, Xiaoya Zhang, Xiaohu Zheng, and Chao Ma

ACM MM 2025 (Oral, CCF A)

arXiv Code Video

@inproceedings{long2025cdupatch,
  title={Cdupatch: Color-driven universal adversarial patch attack for dual-modal visible-infrared detectors},
  author={Long, Jiahuan and Yao, Wen and Jiang, Tingsong and Hou, Jiacheng and Jia, Shuai and Wu, Junqi and Zhang, Xiaoya and Zheng, Xiaohu and Ma, Chao},
  booktitle={Proceedings of the 33rd ACM International Conference on Multimedia (ACM MM)},
  year={2025}
}

CDUPatch RGB-to-infrared adaptation and cross-modal training framework

Robust SAM: On the Adversarial Robustness of Vision Foundation Models

Jiahuan Long, Zhengqin Xu, Tingsong Jiang, Wen Yao, Shuai Jia, Chao Ma, and Xiaoqian Chen

AAAI 2025 (CCF A)

arXiv Code Video

@inproceedings{long2025robustsam,
  title={Robust SAM: On the Adversarial Robustness of Vision Foundation Models},
  author={Long, Jiahuan and Xu, Zhengqin and Jiang, Tingsong and Yao, Wen and Jia, Shuai and Ma, Chao and Chen, Xiaoqian},
  booktitle={Proceedings of the AAAI Conference on Artificial Intelligence (AAAI)},
  year={2025}
}

PapMOT: Exploring Adversarial Patch Attack against Multiple Object Tracking

Jiahuan Long, Tingsong Jiang, Wen Yao, Shuai Jia, Weijia Zhang, Weien Zhou, Chao Ma, and Xiaoqian Chen

ECCV 2024 (CSRankings)

arXiv Video

@inproceedings{long2024papmot,
  title={PapMOT: Exploring Adversarial Patch Attack against Multiple Object Tracking},
  author={Long, Jiahuan and Jiang, Tingsong and Yao, Wen and Jia, Shuai and Zhang, Weijia and Zhou, Weien and Ma, Chao and Chen, Xiaoqian},
  booktitle={European Conference on Computer Vision (ECCV)},
  year={2024}
}

Parameter-Free Fine-tuning via Redundancy Elimination for Vision Foundation Models

Jiahuan Long, Tingsong Jiang, Wen Yao, Yizhe Xiong, Zhengqin Xu, Shuai Jia, Hanqing Liu, and Chao Ma

AAAI 2026 (CCF A)

arXiv Code

@inproceedings{long2026parameterfree,
  title={Parameter-Free Fine-tuning via Redundancy Elimination for Vision Foundation Models},
  author={Long, Jiahuan and Jiang, Tingsong and Yao, Wen and Xiong, Yizhe and Xu, Zhengqin and Jia, Shuai and Liu, Hanqing and Ma, Chao},
  booktitle={Proceedings of the AAAI Conference on Artificial Intelligence (AAAI)},
  year={2026}
}

Parameter-free channel replacement search pipeline

When Surfaces Lie: Exploiting Wrinkle-Induced Attention Shift to Attack Vision-Language Models

Chengyin Hu, Xuemeng Sun, Jiaju Han, Qike Zhang, Xiang Chen, Xin Wang, Yiwei Wei, and Jiahuan Long^✉

ACM MM 2026 (CCF A)

arXiv

@inproceedings{hu2026surfaces,
  title={When Surfaces Lie: Exploiting Wrinkle-Induced Attention Shift to Attack Vision-Language Models},
  author={Hu, Chengyin and Sun, Xuemeng and Han, Jiaju and Zhang, Qike and Chen, Xiang and Wang, Xin and Wei, Yiwei and Long, Jiahuan},
  booktitle={Proceedings of the 34th ACM International Conference on Multimedia (ACM MM)},
  year={2026}
}

Wrinkle-based structural attack framework

Eva-VLA: Evaluating Vision-Language-Action Models' Robustness Under Real-World Physical Variations

Hanqing Liu, Shouwei Ruan, Jiahuan Long, Junqi Wu, Jiacheng Hou, Huili Tang, Tingsong Jiang, Weien Zhou, and Wen Yao

arXiv, 2025

arXiv Video

@article{liu2025evavla,
  title={Eva-VLA: Evaluating Vision-Language-Action Models' Robustness Under Real-World Physical Variations},
  author={Liu, Hanqing and Ruan, Shouwei and Long, Jiahuan and Wu, Junqi and Hou, Jiacheng and Tang, Huili and Jiang, Tingsong and Zhou, Weien and Yao, Wen},
  journal={arXiv preprint arXiv:2509.18953},
  year={2025}
}

Overview of the Eva-VLA robustness evaluation framework

AdvTiles: Physical Adversarial Camouflage Clothing against Person Detectors via Learnable Tiles

Jinlei Wang^*, Jiahuan Long^*, Mingkai Sun, Yafei Guo, Yuanhao Huang, Ming Wang, Junqi Wu, Jiacheng Hou, Hongbo Chen, Xingxing Wei, Tingsong Jiang, and Wen Yao

arXiv, 2026

arXiv Video

@article{wang2026advtiles,
  title={AdvTiles: Physical Adversarial Camouflage Clothing against Person Detectors via Learnable Tiles},
  author={Wang, Jinlei and Long, Jiahuan and Sun, Mingkai and Guo, Yafei and Huang, Yuanhao and Wang, Ming and Wu, Junqi and Hou, Jiacheng and Chen, Hongbo and Wei, Xingxing and Jiang, Tingsong and Yao, Wen},
  journal={arXiv preprint arXiv:2608.06801},
  year={2026}
}

AdvTiles differentiable camouflage and physical rendering framework

Awards

Outstanding Graduate, Shanghai Jiao Tong University
BYD Scholarship, Shanghai Jiao Tong University
Master of Science with Distinction, University of Edinburgh