WorldDirector：使用持久動態記憶體建構可控世界模擬器

WorldDirector: Building Controllable World Simulators with Persistent Dynamic Memory

July 2, 2026

作者: Hanlin Wang, Hao Ouyang, Qiuyu Wang, Wen Wang, Qingyan Bai, Ka Leong Cheng, Yue Yu, Yixuan Li, Yihao Meng, Zichen Liu, Yanhong Zeng, Yujun Shen, Qifeng Chen

cs.AI

摘要

我們提出WorldDirector，一個高度可控的視頻世界模型框架，專為持久動態物體記憶與無限制視角探索而設計。不同於現有世界模型將物理動態與像素渲染糾纏在一起，並依賴連續視覺觀察來維持運動，我們的框架明確地將語義運動編排與視覺生成解耦。通過利用大語言模型協調3D軌跡與攝影機移動，隨後將這些編排後的軌跡作為視頻生成的控制信號，我們的方法確保了嚴格的物理邏輯與外觀穩定性，成功保留了動態實體的準確視覺身份，即使它們在長時間離開視野後重新進入場景。實驗結果表明，我們的方法能夠以前所未有的可控性與持久動態物體記憶，合成複雜且延伸的事件。專案頁面：https://worlddirector.github.io/

English

We present WorldDirector, a highly controllable video world model framework designed for persistent dynamic object memory and unrestricted viewpoint exploration. Unlike existing world models that entangle physical dynamics with pixel rendering and rely on continuous visual observation to sustain motion, our framework explicitly decouples semantic motion orchestration from visual generation. By leveraging an LLM to coordinate 3D trajectories with camera movements and subsequently employing these orchestrated trajectories as control signals for video generation, our approach ensures strict physical logic and appearance stability, successfully preserving the exact visual identities of dynamic entities even when they re-enter the scene after prolonged periods out of view. Experimental results demonstrate that our method supports the synthesis of complex and extended events with unprecedented controllability and persistent dynamic object memory. Project Page: https://worlddirector.github.io/