Yu Cao

STRUCT Group · Peking University

Yu (Calvin) Cao

I am an undergraduate student in Intelligence Science and Technology at Peking University, advised by Prof. Jiaying Liu in the STRUCT group at the Wangxuan Institute of Computer Technology. My research interests mainly lie in low-level vision tasks, including image compression, deraining, super-resolution, and generative coding.

Email GitHub
Yu (Calvin) Cao

About Me

I study computer vision and enjoy building tools that are useful in practice.

Education

  • B.S. student, School of Electronics Engineering and Computer Science, Peking University

  • High School Attached to Northeast Normal University

Honors and Awards

  • 2024-2025

    National Scholarship

  • 2023-2025

    Peking University Merit Student, awarded for two consecutive academic years

  • 2023-2024

    Peking University BYD Scholarship

  • 2024-2025

    Third Prize, Peking University Jiukun Cup Programming Contest

  • 2023-2024

    Peking University Freshman Third-Class Scholarship

  • 2023-2024

    Outstanding Admissions Volunteer, Peking University

Research

My current work focuses on representation, restoration, and compression for visual signals.

I am a member of the STRUCT group, which studies theoretical foundations and practical applications of image/video processing, multimedia communications, and computer vision. My current interests focus on low-level vision tasks where image quality, compact representation, and semantic understanding meet. Recent directions include image compression, video deraining, super-resolution, and generative coding with semantic guidance.

Image Compression Deraining Super-Resolution Generative Coding Low-Level Vision

Publications and Patents

Accepted paper, patent applications, and standardization proposals.

Paper

  1. Haowei Kuang, Yu Cao, Wenhan Yang, Zongming Guo and Jiaying Liu. "Connecting Generation with Compression: Unified Generative Coding with Semantic Guidance", Accepted by IEEE ICME, 2026.

PCT Patent Application

  1. 一种基于掩码补偿的跨粒度在线图像压缩方法及系统。 专利申请人：刘家瑛、匡浩玮、曹彧、郭宗明。专利申请号：PCT/CN2025/135826，申请时间：2025年11月。

Chinese Patent Applications

  1. 一种基于掩码补偿的跨粒度在线图像压缩方法及系统。 专利申请人：刘家瑛、匡浩玮、曹彧、郭宗明。专利申请号：202511168749.1，申请时间：2025年8月。
  2. 一种基于内容-外观解耦的图像簇压缩方法及系统。 专利申请人：刘家瑛、曹彧、匡浩玮、杨文瀚。专利申请号：202610707025.8，申请时间：2026年5月。
  3. 一种基于持续学习与自监督学习的视频去雨方法及系统。 专利申请人：刘家瑛、张语菲、刘明昊、曹彧、杨文瀚。专利申请号：202610852175.8，申请时间：2026年6月。
  4. 一种基于语义引导的统一生成式图像压缩方法及系统。 专利申请人：刘家瑛、董一凡、匡浩玮、曹彧、杨文瀚。专利申请号：202610835226.6，申请时间：2026年6月。

Standardization Proposals

  1. AITISA (AI M2382)：基于视频-特征联合编码的高效适配编码框架技术进展。 杨文瀚，曹彧，孙少凡，刘家瑛，段凌宇。
  2. AITISA (AI M2377)：面向大模型的多用途视频编码-参考软件开发进展。 冯冬辉，高博，凌清悦，程正雪，宋利，王琪，谷群山，胡洪卫，张乐天，曹彧，孙少凡，刘家瑛，段凌宇，杨文瀚，司航。

Contact

I am happy to discuss research, coursework, coding, and music.

calvincao@stu.pku.edu.cn CalvinXiaocao