Saved in:
Bibliographic Details
Main Authors: Hong, Susung, Karras, Johanna, Martin-Brualla, Ricardo, Kemelmacher-Shlizerman, Ira
Format: Preprint
Published: 2024
Subjects:
Online Access:https://arxiv.org/abs/2412.05279
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866916657651253248
author Hong, Susung
Karras, Johanna
Martin-Brualla, Ricardo
Kemelmacher-Shlizerman, Ira
author_facet Hong, Susung
Karras, Johanna
Martin-Brualla, Ricardo
Kemelmacher-Shlizerman, Ira
contents Recent advancements in text-based diffusion models have accelerated progress in 3D reconstruction and text-based 3D editing. Although existing 3D editing methods excel at modifying color, texture, and style, they struggle with extensive geometric or appearance changes, thus limiting their applications. To this end, we propose Perturb-and-Revise, which makes possible a variety of NeRF editing. First, we perturb the NeRF parameters with random initializations to create a versatile initialization. The level of perturbation is determined automatically through analysis of the local loss landscape. Then, we revise the edited NeRF via generative trajectories. Combined with the generative process, we impose identity-preserving gradients to refine the edited NeRF. Extensive experiments demonstrate that Perturb-and-Revise facilitates flexible, effective, and consistent editing of color, appearance, and geometry in 3D. For 360° results, please visit our project page: https://susunghong.github.io/Perturb-and-Revise.
format Preprint
id arxiv_https___arxiv_org_abs_2412_05279
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Perturb-and-Revise: Flexible 3D Editing with Generative Trajectories
Hong, Susung
Karras, Johanna
Martin-Brualla, Ricardo
Kemelmacher-Shlizerman, Ira
Computer Vision and Pattern Recognition
Recent advancements in text-based diffusion models have accelerated progress in 3D reconstruction and text-based 3D editing. Although existing 3D editing methods excel at modifying color, texture, and style, they struggle with extensive geometric or appearance changes, thus limiting their applications. To this end, we propose Perturb-and-Revise, which makes possible a variety of NeRF editing. First, we perturb the NeRF parameters with random initializations to create a versatile initialization. The level of perturbation is determined automatically through analysis of the local loss landscape. Then, we revise the edited NeRF via generative trajectories. Combined with the generative process, we impose identity-preserving gradients to refine the edited NeRF. Extensive experiments demonstrate that Perturb-and-Revise facilitates flexible, effective, and consistent editing of color, appearance, and geometry in 3D. For 360° results, please visit our project page: https://susunghong.github.io/Perturb-and-Revise.
title Perturb-and-Revise: Flexible 3D Editing with Generative Trajectories
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2412.05279