DreamEditor: Text-Driven 3D Scene Editing with Neural Fields

TL;DR

DreamEditor is a novel framework that enables precise, text-guided editing of neural fields representing 3D scenes, allowing localized modifications with high realism and semantic accuracy.

Contribution

It introduces a mesh-based neural field representation combined with text-guided region identification and optimization for controlled 3D scene editing.

Findings

Accurately edits neural fields based on text prompts

Generates highly realistic textures and geometry

Outperforms previous methods in evaluations

Abstract

Neural fields have achieved impressive advancements in view synthesis and scene reconstruction. However, editing these neural fields remains challenging due to the implicit encoding of geometry and texture information. In this paper, we propose DreamEditor, a novel framework that enables users to perform controlled editing of neural fields using text prompts. By representing scenes as mesh-based neural fields, DreamEditor allows localized editing within specific regions. DreamEditor utilizes the text encoder of a pretrained text-to-Image diffusion model to automatically identify the regions to be edited based on the semantics of the text prompts. Subsequently, DreamEditor optimizes the editing region and aligns its geometry and texture with the text prompts through score distillation sampling [29]. Extensive experiments have demonstrated that DreamEditor can accurately edit neural fields of real-world scenes according to the given text prompts while ensuring consistency in irrelevant areas. DreamEditor generates highly realistic textures and geometry, significantly surpassing previous works in both quantitative and qualitative evaluations.