Zixuan Pan

I am a Ph.D. candidate in Computer Science and Engineering at the University of Notre Dame, advised by Prof. Yiyu Shi. I also work closely with Dr. Jianxu Chen (ISAS). I received my B.Eng. from Zhejiang University in 2022.

My research centers on post-training, representation learning, generative AI, and agentic AI, with applications in medical and microscopy image analysis.

Education

• 2022 – 2026 (exp.) Ph.D. in Computer Science and Engineering, University of Notre Dame
• 2018 – 2022 B.Eng. in Information Engineering, Zhejiang University

Publications

Conference

• SGI: Structured 2D Gaussians for efficient and compact large image representation Zixuan Pan*, Kaiyuan Tang*, Jun Xia, Yifan Qin, Lin Gu, Chaoli Wang, Jianxu Chen, Yiyu Shi CVPR 2026 (Highlight). [Paper] [arXiv] [Code]
• H-CNN-ViT: A hierarchical gated attention multi-branch model for bladder cancer recurrence prediction Xueyang Li, Zongren Wang, Yuliang Zhang, Zixuan Pan, Yu-Jen Chen, Nishchal Sapkota, Gelei Xu, Danny Chen, Yiyu Shi BIBM 2025. [Paper] [arXiv]
• Multi-exit class activation map guided feature masking for unsupervised out-of-distribution detection in medical imaging Zixuan Pan, Yu-Jen Chen, Jun Xia, Max Ficco, Jianxu Chen, Tsung-Yi Ho, Yiyu Shi BIBM 2025. [Paper] [Code]
• Rethinking medical anomaly detection in brain MRI: An image quality assessment perspective Zixuan Pan, Jun Xia, Zheyu Yan, Guoyue Xu, Yifan Qin, Xueyang Li, Yawen Wu, Zhenge Jia, Jianxu Chen, Yiyu Shi BIBM 2025. [Paper] [arXiv] [Website] [Code]
• TSB: Tiny shared block for efficient DNN deployment on NVCIM accelerators Yifan Qin, Zheyu Yan, Zixuan Pan, Wujie Wen, Xiaobo Sharon Hu, Yiyu Shi ICCAD 2024. [Paper] [arXiv]

Journal

• NeFT: Negative feedback training to improve robustness of compute-in-memory DNN accelerators Yifan Qin, Zheyu Yan, Dailin Gan, Jun Xia, Zixuan Pan, Wujie Wen, Xiaobo Sharon Hu, Yiyu Shi IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems (TCAD), 45(2):983–997, 2025. [Paper] [arXiv]
• AutoQC-Bench: A diffusion model and benchmark for automatic quality control in high-throughput microscopy Zixuan Pan*, Justin Sonneck*, Dennis Nagel, Anja Hasenberg, Matthias Gunzer, Yiyu Shi, Jianxu Chen npj Imaging, 3(1):57, 2025. [Paper] [Code] [Data]
• Partial unbalanced feature transport for cross-modality cardiac image segmentation Shunjie Dong, Zixuan Pan, Yu Fu, Dongwei Xu, Kuangyu Shi, Qianqian Yang, Yiyu Shi, Cheng Zhuo IEEE Transactions on Medical Imaging (TMI), 42(6):1758–1773, 2023. [Paper]
• DeU-Net 2.0: Enhanced deformable U-Net for 3D cardiac cine MRI segmentation Shunjie Dong, Zixuan Pan, Yu Fu, Qianqian Yang, Yuanxue Gao, Tianbai Yu, Yiyu Shi, Cheng Zhuo Medical Image Analysis (MedIA), 78:102389, 2022. [Paper]

Preprints

• BIABench: Evaluating AI agents on real-world bioimage analysis tasks Zixuan Pan*, Davide Panzeri*, Lukas Johanns, Marilin Moor, Yu Zhou, Hedi Peterson, Yiyu Shi, Jianxu Chen Under review, 2026 arXiv:2609.34274, 2026. [arXiv] [Website] [Code] [Data]
• Agentic-J: An AI agent for biological microscopy image analysis Lukas Johanns, Marilin Moor, Davide Panzeri, Yu Zhou, Xinyi Chen, Nora Pauly, Zixuan Pan, Matthias Gunzer, Andreas Müller, Yiyu Shi, Hedi Peterson, Jianxu Chen Under review, 2026 arXiv:2606.02080, 2026. [arXiv] [Code]
• Masked diffusion as self-supervised representation learner Zixuan Pan, Jianxu Chen, Yiyu Shi arXiv:2308.05695, 2023. [arXiv] [Code]

* Equal contribution

Work Experience

• Oct 2025 – May 2026 Research Scientist Intern, Canva (Austin, TX) VLM judges with RAG for design quality evaluation; fine-grained reward modeling; RL-based post-training to align design generation with human preference.

Awards & Honors

• 2026Graduate School Professional Development Award, University of Notre Dame
• 2022CSE Select Fellowship, University of Notre Dame
• 2022Outstanding Bachelor Graduate & Outstanding Bachelor Thesis, Zhejiang University
• 2021Meritorious Winner (First Prize), Interdisciplinary Contest in Modeling (ICM)
• 2019 – 2021Zhejiang University Scholarships

Service

Reviewer for NeurIPS, KDD, ICLR, AAAI, ISBI, and Scientific Reports (Nature).