Coupled Diffusion Sampling for Training-Free Multi-View Image Editing

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Alzayer, Hadi, Zhang, Yunzhi, Geng, Chen, Huang, Jia-Bin, Wu, Jiajun
Format: Preprint
Veröffentlicht: 2025
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866918162000248832
author Alzayer, Hadi
Zhang, Yunzhi
Geng, Chen
Huang, Jia-Bin
Wu, Jiajun
author_facet Alzayer, Hadi
Zhang, Yunzhi
Geng, Chen
Huang, Jia-Bin
Wu, Jiajun
contents We present an inference-time diffusion sampling method to perform multi-view consistent image editing using pre-trained 2D image editing models. These models can independently produce high-quality edits for each image in a set of multi-view images of a 3D scene or object, but they do not maintain consistency across views. Existing approaches typically address this by optimizing over explicit 3D representations, but they suffer from a lengthy optimization process and instability under sparse view settings. We propose an implicit 3D regularization approach by constraining the generated 2D image sequences to adhere to a pre-trained multi-view image distribution. This is achieved through coupled diffusion sampling, a simple diffusion sampling technique that concurrently samples two trajectories from both a multi-view image distribution and a 2D edited image distribution, using a coupling term to enforce the multi-view consistency among the generated images. We validate the effectiveness and generality of this framework on three distinct multi-view image editing tasks, demonstrating its applicability across various model architectures and highlighting its potential as a general solution for multi-view consistent editing.
format Preprint
id arxiv_https___arxiv_org_abs_2510_14981
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Coupled Diffusion Sampling for Training-Free Multi-View Image Editing
Alzayer, Hadi
Zhang, Yunzhi
Geng, Chen
Huang, Jia-Bin
Wu, Jiajun
Computer Vision and Pattern Recognition
Artificial Intelligence
We present an inference-time diffusion sampling method to perform multi-view consistent image editing using pre-trained 2D image editing models. These models can independently produce high-quality edits for each image in a set of multi-view images of a 3D scene or object, but they do not maintain consistency across views. Existing approaches typically address this by optimizing over explicit 3D representations, but they suffer from a lengthy optimization process and instability under sparse view settings. We propose an implicit 3D regularization approach by constraining the generated 2D image sequences to adhere to a pre-trained multi-view image distribution. This is achieved through coupled diffusion sampling, a simple diffusion sampling technique that concurrently samples two trajectories from both a multi-view image distribution and a 2D edited image distribution, using a coupling term to enforce the multi-view consistency among the generated images. We validate the effectiveness and generality of this framework on three distinct multi-view image editing tasks, demonstrating its applicability across various model architectures and highlighting its potential as a general solution for multi-view consistent editing.
title Coupled Diffusion Sampling for Training-Free Multi-View Image Editing
topic Computer Vision and Pattern Recognition
Artificial Intelligence
url https://arxiv.org/abs/2510.14981