WorldDirector: Building Controllable World Simulators with Persistent Dynamic Memory

Hugging Face Daily Papers Papers
video-generation world-model controllable-simulation persistent-memory llm 3d-trajectories

Summary

WorldDirector introduces a framework for controllable video world simulation that decouples semantic motion planning from visual rendering using an LLM to coordinate 3D trajectories and camera movements, enabling persistent object memory and stable appearance even after objects re-enter the scene.

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/
Original Article
View Cached Full Text

Cached at: 07/03/26, 03:52 AM

Paper page - WorldDirector: Building Controllable World Simulators with Persistent Dynamic Memory

Source: https://huggingface.co/papers/2607.02517 Authors:

,

,

,

,

,

,

,

,

,

,

,

Abstract

WorldDirector enables controllable video generation with persistent object memory by decoupling semantic motion planning from visual rendering through LLM coordination of 3D trajectories and camera movements.

We present WorldDirector, a highly controllablevideo world modelframework designed forpersistent dynamic object memoryand unrestrictedviewpoint exploration. Unlike existing world models that entangle physical dynamics with pixel rendering and rely on continuous visual observation to sustain motion, our framework explicitly decouplessemantic motion orchestrationfromvisual generation. By leveraging anLLMto coordinate3D trajectorieswithcamera movementsand subsequently employing these orchestrated trajectories as control signals for video generation, our approach ensures strictphysical logicandappearance 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 andpersistent dynamic object memory. Project Page: https://worlddirector.github.io/

View arXiv pageView PDFAdd to collection

Get this paper in your agent:

hf papers read 2607\.02517

Don’t have the latest CLI?curl \-LsSf https://hf\.co/cli/install\.sh \| bash

Models citing this paper0

No model linking this paper

Cite arxiv.org/abs/2607.02517 in a model README.md to link it from this page.

Datasets citing this paper0

No dataset linking this paper

Cite arxiv.org/abs/2607.02517 in a dataset README.md to link it from this page.

Spaces citing this paper0

No Space linking this paper

Cite arxiv.org/abs/2607.02517 in a Space README.md to link it from this page.

Collections including this paper0

No Collection including this paper

Add this paper to acollectionto link it from this page.

Similar Articles

ReWorld: An Interactive World Model with Long-Horizon Memory

Hugging Face Daily Papers

ReWorld presents an interactive world model that decouples short-horizon control from long-horizon memory using mixed attention windows and distribution-matching LoRA distillation, enabling real-time video generation with high action fidelity and recall.

Wonder: Video World Model Done Better

Hugging Face Daily Papers

Wonder is a general-purpose video world model that enables real-time, camera-controllable world exploration from an image or conditional video. It introduces camera conditioning via dense coordinate fields, a sparse attention memory mechanism, and techniques to improve distillation, allowing minute-scale video generation at 16 FPS.

Addressable Memory for Video World Models

Hugging Face Daily Papers

This paper introduces WorldTrace, a training-free memory framework for long-horizon video world models that keeps compressed cache addressable, plus LoopBench, a benchmark for episodic recall after long detours. It improves temporal consistency by +15.5% and episodic recall by +19.5% on LoopBench.