Zijie Meng

Zijie Meng

Peking University / Kuaishou Kling Team

🤓 Greetings!

I am Zijie Meng (孟子杰), a second-year Master’s student at Peking University (PKU). Currently, I am a Research Intern at Kuaishou Kling Team, working with [Jiwen Liu, and Pengfei Wan on the next generation of video foundation models.

I obtained my Double Bachelor’s Degrees in Artificial Intelligence and Finance (from the Wang Yanan Institute for Studies in Economics, WISE) from Xiamen University in 2024. My research journey in Generative AI began at the MAC Lab, supervised by Prof. Rongrong Ji. Since then, I have been fortunate to hone my skills through research internships at ByteDance (Seed Team), 360 AI Research, and Shanghai AI Lab (OpenMMLab).

I am incredibly honored to be the recipient of the Peking University May Fourth Scholarship (the highest individual honor at PKU) and the Xiamen University Golden Kapok Medal (one of the highest honor at XMU, Top 0.01% school-wide).

🚀 Research Vision: Towards the World Model

Since I first encountered Diffusion Models in 2023, I have been captivated by their immense generative power. I believe that generative capability is the ultimate goal for achieving AGI. In my philosophy, achieving “controllable generation at will” is the definitive path to constructing a true World Model. To perfectly control the generation of a world, the model must inherently possess a profound and comprehensive understanding of that world’s physics, semantics, and dynamics.

My representative projects include OmniDirector (Developed on Kling Omni, Camera Control), Kling 3.0 (Subject-ID & Motion Control), ARGUS(ID Control), Orpaint (Visual Mamba-based Inpainting), and Make-a-Game (Game Video Generation).

🔍 Research Interests

Currently, I am focusing on building the next generation of vision intelligence:

  • 1️⃣ AIGC & Video Generation: Focusing on Controllable Video Synthesis, Multi-modal Generation, and integrating 3D Spatial Priors into generative paradigms to build robust World Models.
  • 2️⃣ Agentic World Models: Exploring how AI Agents can assist in constructing and navigating simulated environments.
  • 3️⃣ Vision-Language Models (VLM): Enhancing Multi-modal Alignment, instruction-following, and human-AI interaction capabilities.

📧 I am always open to academic collaborations or discussions regarding Video Foundations and AGI. If you’re interested in my work or seeking research synergy, please feel free to reach me at ymlf@stu.pku.edu.cn.

🔥 News

  • 2026-06 : 🎬 OmniDirector is officially launched on Kling Omni! It enables General Multi-Shot Camera Cloning. Check out our Project Page and YouTube Demo! 🚀🌍

  • 2026-05 : 🪄 Our Motion Controllable Video Generation project has been successfully integrated into Kling 3.0, empowering users with advanced movement control. 🏃‍♂️💨

  • 2026-03 : 🆔 The Subject-ID Injection module is now online as a core feature of the Kling 3.0 ID Subject Library, achieving SOTA identity consistency. 🎭✨

  • 2026-02 : 📑 ARGUS and 3D-RAD are accepted to NeurIPS 2026. Make a Game is accepted to ICASSP 2026. Cheers! 🍻

  • 2025-12 : 📄 Finished the technical report for Avatar 2.0 Identity Preservation, bridging the gap between static ID and dynamic video. 🤖📑

  • 2025-11 : 👷 I start my internship at Kuaishou Kling, focusing on Video Foundation Models. 📸🎬

  • 2025-10 : 📑 Robust Sand Removal is accepted to ACM Multimedia (ACM MM) 2025. Congrats to the team! 🏜️🔍

  • 2025-09 : 🚀 Joined Fan Hua Tech as the AI Team Lead, reconstructing business logic with Generative AI and Agentic workflows. 🛠️🧠

  • 2025-05 : 👑 I am honored to receive the May Fourth Scholarship (北京大学五四奖学金), the Highest Individual Honor at Peking University (Top 0.1%). 🎖️🏛️

  • 2025-02 : 👷 I join ByteDance - Seed Team as a Research Intern, working on high-efficiency video compression and controllable generation. 🎥⚡

  • 2024-09 : 🏫 Officially started my Master’s journey at Peking University. Meanwhile, I join 360 AI Research to explore Flux-based architectures. 🖋️🎓

  • 2024-06 : 🎓 Graduated from Xiamen University with a Double Degree in Artificial Intelligence and Finance (from WISE). Honored as Outstanding Graduate and Outstanding Thesis! 📜🏅

  • 2024-03 : 📑 Orpaint is accepted to Science China Information Sciences (SCIS, CCF-A). 🎨🏛️

  • 2023-07 : 🥇 Admitted to Peking University via Summer Camp, ranking 2/91 in the Department of AI at XMU. 🏔️☀️

  • 2023-06 : 👷 Started my internship at Shanghai AI Lab (OpenMMLab), contributing to MMDetection and MMSegmentation. 💻⭐

  • 2022-11 : 🏆 Won the First Prize in Fujian Mobile Application Innovation Competition for mural protection work. 🏮🎨

  • 2022-08 : 🏆 Awarded First Prize in China-US Young Maker Competition (South China Region). 🛠️🇺🇸

📝 Publications

A full publication list is available on my Google Scholar page.

(*: Equal contribution; †: Corresponding authors.)

🎬 Video Generation, World Model & Multimodal Model

arXiv 2026
OmniDirector

[arXiv 2026 Kling Team] OmniDirector: General Multi-Shot Camera Cloning without Cross-Paired Data
Jiwen Liu, Shuo Li, Zhang Fang, Xiang Li, Yitong Zhou, Zijie Meng, et al.

  • We propose OmniDirector, a general framework for multi-shot camera cloning that operates without the need for cross-paired data, significantly advancing camera control in video synthesis.
arXiv 2026
ARGUS

[arXiv 2026] ARGUS: Stacked Multi-View Identity Mosaic Injection for Subject-Preserving Video Generation
Zijie Meng, Jiwen Liu, Yu Liu, Chen Tong, Xiao Liu, Yingya Zhang, Yong Xu, Pengfei Wan.
[Code] (Internal/Coming soon)

  • We propose ARGUS, a novel framework for subject-preserving video generation using stacked multi-view identity mosaic injection, ensuring high fidelity and temporal consistency.
  • This work was conducted during my internship at Kuaishou Kling, focusing on controllable identity injection in foundation video models.
ICASSP 2026
Make A Game

[ICASSP 2026 oral] Make a Game: A Novel Paradigm for Interactive Game Rendering
Zijie Meng, Jian Che, Bo Wei, Xuesong Cao.
[Award] First Prize in PKU Challenge Cup

  • We introduce a novel paradigm for interactive game rendering using unified tokens and lightweight plugins, enhancing controllability in video generation.
  • Successfully generalized to complex interactive game scenarios, providing a bridge between generative AI and real-time game engines.

Dataset

NeurIPS 2026
3D-RAD

[NeurIPS 2026] 3d-rad: A Comprehensive 3d Radiology Med-vqa Dataset with Multi-temporal Analysis and Diverse Diagnostic Tasks
X Gai, J Liu, Y Li, Zijie Meng, J Wu, Z Liu.

  • We introduce 3D-RAD, the most comprehensive 3D radiology dataset for Medical VQA, supporting multi-temporal analysis and diverse clinical diagnostic tasks.

🎨 Image-Generation & Restoration & Segmentation

SCIS (CCF-A)
Orpaint

[Science China Info. Sci. 2025] Orpaint: A Zero-Shot Inpainting Model for Oracle Bone Inscription Rubbings with Visual Mamba Block
Zijie Meng, Yuer Zeng, Xiang Chang, Tianyang Xu, Fei Chao, Xuesong Cao, Chun Chen, Qiang Shen.
[Journal] JCR-Q1, CCF-A

  • We propose Orpaint, the first zero-shot inpainting model specifically designed for Oracle Bone Inscription (甲骨文) restoration.
  • By integrating the Visual Mamba Block into the Diffusion denoising network, we achieve significantly faster inference and better structural restoration for damaged ancient rubbings.
ACM MM 2025
Sand Removal

[ACM MM 2025] Robust Single Image Sand Removal by Leveraging Uncertainty-aware SAM Priors and Prompt Learning with Refined Perceptual Loss
Bo Wei, Huafeng Liu, Cheng Qian, Zizheng Li, Wenbo Wu, Zijie Meng.
CCF-A Conference

  • We address the challenging task of sand-dust image restoration by leveraging uncertainty-aware SAM (Segment Anything Model) priors and prompt learning.
  • My contribution focused on the Llama3 fine-tuning for generating refined perceptual instructions.
ICME 2026
MST-CLIPIQA

[ICME 2026 Spotlight] Decoupling Semantics from Distortions: Multi-Scale Two-Stream Vision-Language Alignment for AI-Generated Image Quality Assessment
Zijie Meng.
CCF-B | GitHub

  • We propose a multi-scale two-stream vision-language alignment framework that decouples semantic understanding from distortion perception for robust AI-generated image quality assessment.
  • My contribution focused on the overall framework design and vision-language alignment strategy.
MICCAI 2025
SynPo

[MICCAI 2025] SynPo: Boosting Training-Free Few-Shot Medical Segmentation via High-Quality Negative Prompts
Y Liu, H Xiao, J Chai, Y Zhang, R Wang, Zijie Meng, Z Luo.
CCF-B Conference / Medical AI Top Conference

  • We propose SynPo, which boosts training-free medical image segmentation by utilizing high-quality negative prompts to refine few-shot boundary detection.

💻 Internships

  • 2025.09 - Present, Kuaishou Kling Team.
    • Advisor: Jiwen Liu and Pengfei Wan.
    • Focus: Controllable Video Foundation Models(ID,Motion,Camera) & World Models.
  • 2025.02 - 2025.07, ByteDance - Seed Team.
    • Advisor: Ziyang Liu and Hang Li (Author of Statistical Learning Methods).
    • Focus: Unified Compression for Video Generation & Controllable Driving Video Synthesis.
  • 2024.09 - 2024.12, 360 AI Research, Visual Engine Department.
  • 2023.11 - 2024.03, (OpenMMLab).
    • Focus: Development and evaluation of MMDetection and MMSegmentation(29.2K+ Stars).

🎓 Academic Service

  • Reviewing
    • Conferences:
      • NeurIPS 2026
      • ICML 2026
      • ACM Multimedia (ACM MM) 2026
      • ICME 2025-2026
      • ICASSP 2025
    • Journals:
      • Pattern Recognition (PR)
      • Knowledge-Based Systems (KBS) (Invited Reviewer)

💬 Miscellaneous

  • Research Taste & Vision:
    I am currently all-in on Diffusion Models and Video Generation. My ultimate research goal is to construct a robust World Model via controllable Video Foundation Models, bridging the gap between generative AI and physical reality.
  • Arts & Sports:
    • 🖌️ Calligraphy: I hold the Level 9 (Highest Professional Grade) Certificate from the China Academy of Art (中国美术学院). I see calligraphy as a way to practice patience and structural thinking.
    • 🏓 Table Tennis: An avid player with a rating of ~1650 on Kaiqiu. I love the fast-paced strategy and physical intuition involved in every stroke.
    • 🎹 Piano: Grade 7 (Amateur) certified by the China Conservatory of Music.
    • 🏅 Athletics: Passionate about High Jump and Swimming. I am a huge fan of various sports including Snooker, Football, and Basketball.
  • Public Speaking & Hosting: I enjoy the stage as much as the lab. It’s an honor to serve as the Host for prestigious events:
    • 2026 Peking University Alumni New Year Gala
    • 2025 PKU School of Software and Microelectronics New Year Gala
  • Globe Explorer:
    Traveling is my way of resetting. I have explored 8 countries including France 🇫🇷, Spain 🇪🇸, Thailand 🇹🇭, Singapore 🇸🇬, Mongolia 🇲🇳, Malaysia 🇲🇾, and the Philippines 🇵🇭. Domestically, I have left my footprints in 109 cities across China. My dream is to explore every corner of the globe. 🌍

🎖 Honors and Awards

Below, I exhaustively list some of my Honors and Awards that inspire me a lot.

  • 2026-05    Bronze Medal in the 3rd National Advanced Computing Technology Competition (Top 1% / Top 10 teams nationwide) (第三届全国先进计算技术大赛全国总决赛铜奖)
  • 2025-05    May Fourth Scholarship of Peking University (The highest individual honor of PKU, Top 0.1% / Only 2 candidates in our college) (北京大学五四奖学金)
  • 2025-10    Merit Student of Peking University (北京大学三好学生)
  • 2024-05    Golden Kapok Medal of Xiamen University (The highest honor of XMU, Top 0.01% / Only 10 students school-wide) (厦门大学金木棉奖章)
  • 2024-06    Outstanding Graduate of Xiamen University (厦门大学优秀毕业生)
  • 2024-06    Outstanding Graduation Thesis of Xiamen University (Top score in Department) (厦门大学优秀毕业设计)
  • 2024-05    ICBC Scholarship (工商银行奖学金)
  • 2024-03    Outstanding Student Leader of Xiamen University (厦门大学优秀学生干部)
  • 2023-10    National Scholarship (Top 1% among all undergraduates) (国家奖学金)
  • 2023-09    First-Class Academic Excellence Scholarship of Xiamen University (Top 3%) (学业优秀一等奖学金)
  • 2023-05    Merit Student of Xiamen University (厦门大学优秀三好学生)
  • 2022-11    First Prize in Fujian Mobile Application Innovation Competition (Provincial Level) (福建省移动应用创新赛省级一等奖)
  • 2022-09    First-Class Academic Excellence Scholarship of Xiamen University (Top 3%) (学业优秀一等奖学金)
  • 2022-08    First Prize in China-US Young Maker Competition (South China Region) (中美青年创客大赛华南赛区一等奖)
  • 2021-11    Second Prize in National Mathematical Modeling Contest (Top 5%) (全国大学生数学建模竞赛二等奖)
  • 2021-09    First-Class Academic Excellence Scholarship of Xiamen University (Top 3%) (学业优秀一等奖学金)
  • 2021-05    Honorable Mention in Mathematical Contest in Modeling (MCM/ICM)