Zongze Du

Incoming Ph.D. Student, Zhejiang University CAD&CG Lab

zongze-du.jpg

Hangzhou, China

3220105581 [at] zju [dot] edu [dot] cn

中文主页

I am an incoming Ph.D. student at Zhejiang University CAD&CG Lab, advised by Prof. Chunhua Shen and Prof. Hao Chen. My research interests lie at the intersection of multimodal reasoning, embodied intelligence, and video generation.

I received my B.E. in Computer Science from Zhejiang University and studied in Chu Kochen Honors College. I am also the founder of AIRA, an AI research community at Zhejiang University, and a core CUDA programmer at ZJUSCT with hands-on experience in multi-GPU optimization and student supercomputing competitions.

My current work focuses on building AI systems that connect frontier research with real products and industry impact, including omnimodal reasoning, high-fidelity video generation, companion robots, and high-performance computing optimization.

news

Jan 01, 2026 FrontierX AURA companion robot was showcased at CES 2026.
Dec 01, 2025 Omni-R1 was accepted to NeurIPS 2025.
Mar 01, 2025 An AIRA team I mentored won 1st Prize in the Ascend LLM track at the Huawei ICT National Finals.

selected publications

  1. CVPR
    Preserving Source Video Realism: High-Fidelity Face Swapping for Cinematic Quality
    Zekai Luo, Zongze Du, Zhouhang Zhu, and 7 more authors
    In IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2026
  2. NeurIPS
    Omni-R1: Reinforcement Learning for Omnimodal Reasoning via Two-System Collaboration
    Hao Zhong, Muzhi Zhu, Zongze Du, and 6 more authors
    In Advances in Neural Information Processing Systems, 2025
    Equal contribution
  3. ICML
    ACTIVE-O3: Empowering MLLMs with Active Perception via Pure Reinforcement Learning
    Muzhi Zhu, Hao Zhong, Canyu Zhao, and 9 more authors
    In International Conference on Machine Learning, 2026