Haoji Zhang (张颢继)

I am a second-year master's student at Shenzhen International Graduate School, Tsinghua University, majoring in Data Science and Information Technology. I expect to graduate in 2027. I am fortunate to be supervised by Prof. Yansong Tang in IVG@SZ group.

Before that, I received my B.S. degree in Mathematics and Physics and B.E. degree in Materials Science and Engineering from Tsinghua University in 2024.

I feel honored to have internships in Prof. Jun Zhu's TSAIL Lab, Prof. Gao Huang's LEAP Lab, ByteDance Seed / Intelligent Creation, working with amazing mentors such as Dr. Yulin Wang, Prof. Xiaojie Jin, Dr. Jiashi Feng, Dr. Longyin Wen.

My research focuses on building efficient and powerful MLLMs , including:

  • Efficient Long Video Understanding: AR, VQA, VTG, STVG
  • Large-scale MLLM Training: PE, PT, SFT, Post-Train (GRPO, Tool-RL)
  • Agentic Model and Multimodal Reasoning
  • Unified Understanding and Generation Model

Google Scholar  /  Email  /  Github  /  LinkedIn

News

  • 2026.01: We introduce Vidi2.5, the open-source SOTA MLLM for long video TR, STG, VQA and plot reasoning (Core author)
  • 2025.11: SC-CLIP is accepted by TIP (CCF-A, Q1, IF=13.7)
  • 2025.11: We introduce STVG-o1, enabling chain-of-bboxes reasoning for spatial-temporal video grounding (Co-First author)
  • 2025.08: We open-source Thinking With Videos / VITAL, the first agentic video reasoning RL framework (First author)
  • 2025.05: Flash-VStream is accepted by ICCV'25 (First author, CCF-A)
  • 2025.05: Ponder & Press is accepted by ACL'25 Findings (Co-First author, CCF-A)
  • 2024.12: Uni-AdaFocus is accepted by TPAMI (Co-First author, CCF-A, Q1, IF=20.4)
  • 2024.06: Flash-VStream wins the 1st place on LOVEU@CVPR'24: Long-term VQA
  • 2023.02: PREIM3D is accepted by CVPR'23
  • Publications and Preprints [ Full List ]
    (* Equal Contribution, Corresponding Author)
    Honors and Awards

  • First-Class Outstanding Student Scholarship, Tsinghua University, 2025. (清华大学校级一等综合优秀奖学金)
  • Beijing Outstanding Bachelor's Graduate, 2024. (北京市优秀毕业生, Top 5% in Tsinghua University)
  • First-Class Outstanding Student Scholarship, Tsinghua University, 2023. (清华大学校级一等综合优秀奖学金)
  • First-Class Outstanding Student Scholarship, Tsinghua University, 2022. (清华大学校级一等综合优秀奖学金)
  • First-Class Outstanding Student Scholarship, Tsinghua University, 2021. (清华大学校级一等综合优秀奖学金)
  • THUWC2019 Gold Medal, 2019. (清华大学全国优秀中学生信息学冬令营金牌)
  • NOIWC2019 Silver Medal, 2019. (第36届全国信息学奥林匹克冬令营银牌)
  • NOI2019 Bronze Medal, 2019. (第36届全国信息学奥林匹克竞赛铜牌)
  • Academic Services

  • Converence Reviewer: ICCV, AAAI, ICLR, CVPR
  • Journal Reviewer: JVCIR
  • Contact

  • Email: hodge013140 [AT] gmail [DOT] com
  • WeChat: the same as email

  • Website Template

    © Haoji Zhang | Last updated: Feb 4, 2026