PrEditor3D: Fast and Precise 3D Shape Editing

TL;DR

PrEditor3D introduces a rapid, training-free method for precise 3D shape editing by combining multi-view image editing with a 3D segmentation pipeline, ensuring high-quality, targeted modifications within minutes.

Contribution

It presents a novel 3D editing framework that aligns multi-view image editing with 3D segmentation, enabling fast, accurate shape modifications without training.

Findings

Achieves high-quality 3D editing within minutes.

Outperforms previous methods in precision and speed.

Preserves unedited regions effectively.

Abstract

We propose a training-free approach to 3D editing that enables the editing of a single shape within a few minutes. The edited 3D mesh aligns well with the prompts, and remains identical for regions that are not intended to be altered. To this end, we first project the 3D object onto 4-view images and perform synchronized multi-view image editing along with user-guided text prompts and user-provided rough masks. However, the targeted regions to be edited are ambiguous due to projection from 3D to 2D. To ensure precise editing only in intended regions, we develop a 3D segmentation pipeline that detects edited areas in 3D space, followed by a merging algorithm to seamlessly integrate edited 3D regions with the original input. Extensive experiments demonstrate the superiority of our method over previous approaches, enabling fast, high-quality editing while preserving unintended regions.