DreamScene: 3D Gaussian-based Text-to-3D Scene Generation via Formation Pattern Sampling
This addresses the problem of generating high-quality, consistent, and editable 3D scenes from text for applications in gaming, film, and architecture, representing a novel method for known bottlenecks.
The paper tackles text-to-3D scene generation by proposing DreamScene, a framework that uses 3D Gaussian-based methods and formation pattern sampling to improve quality, consistency, and editing flexibility, achieving superiority over state-of-the-art techniques in experiments.
Text-to-3D scene generation holds immense potential for the gaming, film, and architecture sectors. Despite significant progress, existing methods struggle with maintaining high quality, consistency, and editing flexibility. In this paper, we propose DreamScene, a 3D Gaussian-based novel text-to-3D scene generation framework, to tackle the aforementioned three challenges mainly via two strategies. First, DreamScene employs Formation Pattern Sampling (FPS), a multi-timestep sampling strategy guided by the formation patterns of 3D objects, to form fast, semantically rich, and high-quality representations. FPS uses 3D Gaussian filtering for optimization stability, and leverages reconstruction techniques to generate plausible textures. Second, DreamScene employs a progressive three-stage camera sampling strategy, specifically designed for both indoor and outdoor settings, to effectively ensure object-environment integration and scene-wide 3D consistency. Last, DreamScene enhances scene editing flexibility by integrating objects and environments, enabling targeted adjustments. Extensive experiments validate DreamScene's superiority over current state-of-the-art techniques, heralding its wide-ranging potential for diverse applications. Code and demos will be released at https://dreamscene-project.github.io .