About Me

Hello! I’m Zishuo Wang (王梓烁), currently a Master’s Student at Wangxuan Institute of Computer Technology, Peking University. My research interests primarily focus on Multimodal and Computer Vision, with a particular emphasis on Efficient Large Vision-Language Models.

I am pursuing my Master’s degree at Multimedia Information Processing Lab (MIPL), under the supervision of Prof. Yuxin Peng.

Beyond academia, I am passionate about football and Hip-Hop music, which helps me stay creative and motivated in my research work.

• Email: wangzishuo@pku.edu.cn
• Github
• Google Scholar

Education

• B.Sc. in Computer Science and Technology, Peking University, Sept. 2019 - Jul. 2023
• M.Sc. in Multimedia Intelligence, Peking University, Sept. 2023 - Now

Publications

• SIA-OVD: Shape-Invariant Adapter for Bridging the Image-Region Gap in Open-Vocabulary Detection Paper Code

  Zishuo Wang, Wenhao Zhou, Jinglin Xu, Yuxin Peng.

  ACM Multimedia (ACM MM ) 2024.

• FineParser: A Fine-grained Spatio-temporal Action Parser for Human-centric Action Quality Assessment Paper Code Project Page

  Jinglin Xu, Sibo Yin, Guohao Zhao, Zishuo Wang, Yuxin Peng.

  IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR Oral ) 2024.

• FinePruner: Unbiased Attention-Head-Level Fine-grained Token Reduction for Efficient Inference of Large Vision-Language Models Paper Code

  Zishuo Wang, Xiangtian Zheng, Yuxin Peng.

  IEEE Transactions on Image Processing (TIP ) 2026.

• A Survey on Fine-Grained Multimodal Large Language Models Paper

  Yuxin Peng, Zishuo Wang, Geng Li, Xiangtian Zheng, Sibo Yin, Hulingxiao He.

  Chinese Journal of Electronics (CJE ) 2026.

• Ctp2Fic: From Coarse-grained Token Pruning to Fine-grained Token Clustering for LVLM Inference Acceleration Paper

  Yulong Lei, Zishuo Wang, Jinglin Xu, Yuxin Peng.

  Elsevier Displays, China National Conference on Multimedia (ChinaMM ) 2025, SSRN’s Top Downloads list.

Thank you for visiting, and I look forward to connecting with you!