NeuSEditor: From Multi-View Images to Text-Guided Neural Surface Edits

Nail Ibrahimli, Julian F. P. Kooij, Liangliang Nan

Keywords: neural rendering + diffusion models + 3D scene editing + implicit fields + multi-view data

Abstract: Implicit surface representations are valued for their compactness and continuity, but they pose significant challenges for editing. Despite recent advancements, existing methods often fail to preserve identity and maintain geometric consistency during editing. To address these challenges, we present NeuSEditor, a novel method for text-guided editing of neural implicit surfaces derived from multi-view images. NeuSEditor introduces an identity-preserving architecture that efficiently separates scenes into foreground and background, enabling precise modifications without altering the scene-specific elements. Our geometry-aware distillation loss significantly enhances rendering and geometric quality. Our method simplifies the editing workflow by eliminating the need for continuous dataset updates and source prompting. NeuSEditor outperforms recent state-of-the-art methods, delivering superior quantitative and qualitative results; in fact, users preferred our pipeline in 28 out of 34 experiments. For more visual results, visit neuseditor.github.io.

Submission Number: 110