# Look-Before-Move: Narrative-Grounded World Visual Attention in Dynamic 3D Story Worlds

arXiv: https://arxiv.org/abs/2606.26964 · HTML: https://mlsanity.com/paper/2606.26964/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Placed in Video-generation world models and simulators, one of 11 directions in World models.
- Builds on 16 of its 38 references that are in the map.
- First posted on arXiv 25 Jun 2026.

## Placed in

- World models / [Video-generation world models and simulators](https://mlsanity.com/wm/video-generation-world-models-and-simulators/)

## Builds on (most cited first, 16 of 16)

- [Qwen2-VL: Enhancing Vision-Language Model's Perception of the World at Any Resolution](https://mlsanity.com/paper/2409.12191/) (2024)
- [Gemini 1.5: Unlocking multimodal understanding across millions of tokens of context](https://mlsanity.com/paper/2403.05530/) (2024)
- [Language Agent Tree Search Unifies Reasoning Acting and Planning in Language Models](https://mlsanity.com/paper/2310.04406/) (2023)
- [CameraCtrl: Enabling Camera Control for Text-to-Video Generation](https://mlsanity.com/paper/2404.02101/) (2024)
- [CamCo: Camera-Controllable 3D-Consistent Image-to-Video Generation](https://mlsanity.com/paper/2406.02509/) (2024)
- [VD3D: Taming Large Video Diffusion Transformers for 3D Camera Control](https://mlsanity.com/paper/2407.12781/) (2024)
- [Towards Understanding Camera Motions in Any Video](https://mlsanity.com/paper/2504.15376/) (2025)
- [SceneWeaver: All-in-One 3D Scene Synthesis with an Extensible and Self-Reflective Agent](https://mlsanity.com/paper/2509.20414/) (2025)
- [FilmAgent: A Multi-Agent Framework for End-to-End Film Automation in Virtual 3D Spaces](https://mlsanity.com/paper/2501.12909/) (2025)
- [Cavia: Camera-controllable Multi-view Video Diffusion with View-Integrated Attention](https://mlsanity.com/paper/2410.10774/) (2024)
- [Dynamic Storyboard Generation in an Engine-based Virtual Environment for Video Production](https://mlsanity.com/paper/2301.12688/) (2023)
- [WorldCraft: Photo-Realistic 3D World Creation and Customization via LLM Agents](https://mlsanity.com/paper/2502.15601/) (2025)
- [PAT3D: Physics-Augmented Text-to-3D Scene Generation](https://mlsanity.com/paper/2511.21978/) (2025)
- [Story3D-Agent: Exploring 3D Storytelling Visualization with Large Language Models](https://mlsanity.com/paper/2408.11801/) (2024)
- [Camera Trajectory Generation: A Comprehensive Survey of Methods, Metrics, and Future Directions](https://mlsanity.com/paper/2506.00974/) (2025)
- [StoryBlender: Inter-Shot Consistent and Editable 3D Storyboard with Spatial-temporal Dynamics](https://mlsanity.com/paper/2604.03315/) (2026)

## Abstract (arXiv)

As embodied AI and world models increasingly operate in dynamic 3D environments, visual perception must move beyond passively interpreting given observations toward actively deciding what to observe. We study this problem through camera planning in dynamic 3D story worlds, where the camera must not only generate smooth motion, but also decide what visual evidence should be acquired before it moves. We formulate this capability as Narrative-Grounded World Visual Attention, where the camera acts as an embodied observer that determines what to observe, how to compose the observation, and how to shift attention over time under narrative intent and physical 3D constraints. To realize this capability, we propose Look-Before-Move, a camera planning framework that separates observation specification from motion execution. It first builds a Semantic Observation Contract to convert directorial intent into executable visual constraints, then performs Monte Carlo Viewpoint Search to find narrative-compliant and geometrically feasible viewpoints, and finally applies Semantic Trajectory Grounding to connect selected viewpoints into continuous, collision-aware, and temporally coherent camera motion. We further construct a dynamic 3D Story World Benchmark based on StoryBlender, covering 50 stories, 457 scenes, and 1585 shots with animated characters, semantic scene configurations, and executable 3D environments. Experiments show that our framework improves subject perception, intent consistency, and trajectory quality over representative baselines, demonstrating the importance of organizing visual attention before generating camera motion.