Fengpeng Li | King Abdullah University of Science and Technology (KAUST)

About Me

I am Kemou Li, currently a second-year PhD student in Computer Science at State Key Laboratory of Internet of Things for Smart City (SKL-IoTSC), University of Macau, advised by Prof. Jiantao Zhou. Prior to that, I earned my BS degree in Mathematics from Sun Yat-sen University and my MS degree in Artificial Intelligence from University of Macau. I also work closely with Dr. Fengpeng Li, Dr. Qizhou Wang, Prof. Haiwei Wu, and Prof. Bo Han.

I am currently focusing on machine unlearning for LLMs, with an emphasis on theoretical foundations. I am also interested in broader topics in trustworthy foundation models, machine learning, and AI security. I am always open to collaboration or exchanging ideas on related topics.

I am looking for visiting PhD opportunity. Please feel free to contact me about any potential openings :)

News

2026-03 Our team wins the 6th place (6/511) at the NTIRE @ CVPR 2026: Robust AI-Generated Image Detection in the Wild Challenge.
2026-02 One paper on image forensics accepted to CVPR 2026 Highlight. See you in Denver!
2026-01 Two papers on LLM unlearning and diffusion model concept erasure accepted to ICLR 2026. See you in Rio de Janeiro!
2025-06 One paper on adversarial training accepted to IEEE TIFS.
2025-06 Pass my PhD qualifying exam.
2025-05 One paper on learning with noisy labels accepted to IJCV.
2025-02 One paper on font watermarking accepted to IEEE TMM.
2025-01 Start remote research at TMLR Group with Dr. Qizhou Wang and Prof. Bo Han.
2024-09 One paper on adversarial training accepted to NeurIPS 2024.
2024-09 Our team wins the championship (1/706) at the Global Multimedia Deepfake Detection Challenge (Image Track) at the 2024 Inclusion Conference on the Bund.
2024-08 Start my PhD journey at the University of Macau.
2023-12 One paper on learning with noisy labels accepted to AAAI 2024 Oral (2.2%).

Selected Publications

* = Equal Contribution

arXiv-26

kNNProxy: Efficient Training-Free Proxy Alignment for Black-Box Zero-Shot LLM-Generated Text Detection

Kahim Wong, Kemou Li, Haiwei Wu, Jiantao Zhou

arXiv preprint, 2026

PDF
CVPR-26 Highlight

Editprint: General Digital Image Forensics via Editing Fingerprint with Self-Augmentation Training

Haiwei Wu*, Kemou Li*, Yuanman Li, Jiantao Zhou

Conference on Computer Vision and Pattern Recognition (CVPR), 2026

Code CCF A Highlight
ICLR-26

LLM Unlearning with LLM Beliefs

Kemou Li, Qizhou Wang, Yue Wang, Fengpeng Li, Jun Liu, Bo Han, Jiantao Zhou

International Conference on Learning Representations (ICLR), 2026

PDF CCF A
ICLR-26

AEGIS: Adversarial Target-Guided Retention-Data-Free Robust Concept Erasure from Diffusion Models

Fengpeng Li, Kemou Li, Qizhou Wang, Bo Han, Jiantao Zhou

International Conference on Learning Representations (ICLR), 2026

PDF CCF A
TIFS-25

Toward Robust Deep Learning via Core Feature-aware Adversarial Training

Fengpeng Li*, Kemou Li*, Haiwei Wu, Jinyu Tian, Jiantao Zhou

IEEE Transactions on Information Forensics and Security (TIFS), 2025

PDF Code CCF A
IJCV-25

RML++: Regroup Median Loss for Combating Label Noise

Fengpeng Li, Kemou Li, Qizhou Wang, Bo Han, Jinyu Tian, Jiantao Zhou

International Journal of Computer Vision (IJCV), 2025

PDF Code CCF A
TMM-25

FontGuard: A Robust Font Watermarking Approach Leveraging Deep Font Knowledge

Kahim Wong, Jicheng Zhou, Kemou Li, Yain-Whar Si, Xiaowei Wu, Jiantao Zhou

IEEE Transactions on Multimedia (TMM), 2025

PDF Code CCF A
NeurIPS-24

DAT: Improving Adversarial Robustness via Generative Amplitude Mix-up in Frequency Domain

Fengpeng Li, Kemou Li, Haiwei Wu, Jinyu Tian, Jiantao Zhou

Conference on Neural Information Processing Systems (NeurIPS), 2024

PDF Code CCF A
AAAI-24 Oral

Regroup Median Loss for Combating Label Noise

Fengpeng Li, Kemou Li, Jinyu Tian, Jiantao Zhou

AAAI Conference on Artificial Intelligence (AAAI), 2024

PDF Code CCF A Oral (2.2%)