Kejia Zhang

Master Student

Xiamen University

About Me

I am a final-year Master student with the School of Information at Xiamen University, fortunate to be advised by Prof. Zhiming Luo. Previously, I received a B.S.c degree in Mathematics from Anhui University, where I worked closely with Prof. Haifeng Zhang during my undergraduate studies.

🎓 Actively seeking Ph.D. opportunities for Fall 2026! If my background and research interests happen to align with the criteria of your group, it would be my great honor to further connect & collaborate with you!

Interests

Trustworthy LVMs
Computer Vision
Neural Network Robustness

Education

Department of Artificial Intelligence, 2023
Xiamen University
School of Mathematical Sciences, 2019
Anhui University

Publications

TARS: MinMax Token-Adaptive Preference Strategy for MLLM Hallucination Reduction

Preprint

Kejia Zhang, Keda Tao, Zhiming Luo, Chang Liu, Jiasheng Tang, Huan Wang

When Tokens Talk Too Much: A Survey of Multimodal Long-Context Token Compression across Images, Videos, and Audios

Preprint

Kele Shao, Keda Tao, Kejia Zhang, Sicheng Feng, Mu Cai, Yuzhang Shang, Haoxuan You, Can Qin, Yang Sui, Huan Wang

When Tokens Talk Too Much: A Survey of Multimodal Long-Context Token Compression across Images, Videos, and Audios

Towards Adversarial Robustness via Debiased High-Confidence Logit Alignment

ICCV 2025

Kejia Zhang, Juanjuan Weng, Shaozi Li, Zhiming Luo

Towards Adversarial Robustness via Debiased High-Confidence Logit Alignment

Mitigating Low-Frequency Bias: Feature Recalibration and Frequency Attention Regularization for Adversarial Robustness

Neural Networks (JCR-Q1, IF=6.3)

Kejia Zhang, Juanjuan Weng, Yuanzheng Cai, Zhiming Luo, Shaozi Li

Poison as Cure: Visual Noise for Mitigating Object Hallucinations in LVMs

Preprint

Kejia Zhang, Keda Tao, Jiasheng Tang, Huan Wang

Harmonizing Feature Maps: A Graph Convolutional Approach for Enhancing Adversarial Robustness

Preprint

Kejia Zhang, Junwei Wu, Juanjuan Weng, Guoqing Yang, Zhiming Luo, Shaozi Li

Extracting Higher Order Topological Semantic via Motif-Based Deep Graph Neural Networks

IEEE TCSS (JCR-Q1, IF=4.9)

Kejia Zhang, Xiao Ding, Bingbing Xiang, Haifeng Zhang, Zhongkui Bao

Extracting Higher Order Topological Semantic via Motif-Based Deep Graph Neural Networks

🏅 Honors & Awards

ICCV BP Travel Award, 2025
National Encouragement Scholarship, 2023
National Second Prize, Chinese Undergraduate Mathematical Contest in Modeling (CUMCM), 2022
Academic Technology Scholarship, 2021 & 2022