GaussCtrl: Multi-View Consistent Text-Driven 3D Gaussian Splatting Editing

TL;DR

GaussCtrl introduces a multi-view consistent, text-driven editing technique for 3D Gaussian Splatting scenes, enabling faster edits with improved visual quality by leveraging depth-conditioned and attention-based latent alignment.

Contribution

The paper presents a novel multi-view consistent editing approach for 3D Gaussian Splatting that outperforms previous methods in speed and visual fidelity.

Findings

Faster editing compared to prior methods.

Higher visual quality in edited 3D scenes.

Effective multi-view consistency maintained during editing.

Abstract

We propose GaussCtrl, a text-driven method to edit a 3D scene reconstructed by the 3D Gaussian Splatting (3DGS). Our method first renders a collection of images by using the 3DGS and edits them by using a pre-trained 2D diffusion model (ControlNet) based on the input prompt, which is then used to optimise the 3D model. Our key contribution is multi-view consistent editing, which enables editing all images together instead of iteratively editing one image while updating the 3D model as in previous works. It leads to faster editing as well as higher visual quality. This is achieved by the two terms: (a) depth-conditioned editing that enforces geometric consistency across multi-view images by leveraging naturally consistent depth maps. (b) attention-based latent code alignment that unifies the appearance of edited images by conditioning their editing to several reference views through self and cross-view attention between images' latent representations. Experiments demonstrate that our method achieves faster editing and better visual results than previous state-of-the-art methods.