Comp4D: LLM-Guided Compositional 4D Scene Generation
This addresses the challenge of 4D scene generation for applications like animation and virtual reality, offering a novel compositional approach to overcome dataset limitations, though it is incremental in building on existing diffusion models.
The paper tackles the problem of generating 4D scenes (3D objects over time) from text prompts by introducing Comp4D, a framework that uses LLMs to decompose prompts into entities and trajectories, then composes scenes with a score distillation technique, resulting in superior visual quality, motion fidelity, and object interactions compared to prior methods.
Recent advancements in diffusion models for 2D and 3D content creation have sparked a surge of interest in generating 4D content. However, the scarcity of 3D scene datasets constrains current methodologies to primarily object-centric generation. To overcome this limitation, we present Comp4D, a novel framework for Compositional 4D Generation. Unlike conventional methods that generate a singular 4D representation of the entire scene, Comp4D innovatively constructs each 4D object within the scene separately. Utilizing Large Language Models (LLMs), the framework begins by decomposing an input text prompt into distinct entities and maps out their trajectories. It then constructs the compositional 4D scene by accurately positioning these objects along their designated paths. To refine the scene, our method employs a compositional score distillation technique guided by the pre-defined trajectories, utilizing pre-trained diffusion models across text-to-image, text-to-video, and text-to-3D domains. Extensive experiments demonstrate our outstanding 4D content creation capability compared to prior arts, showcasing superior visual quality, motion fidelity, and enhanced object interactions.