PCDreamer: Point Cloud Completion Through Multi-view Diffusion Priors

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Wei, Guangshun, Feng, Yuan, Ma, Long, Wang, Chen, Zhou, Yuanfeng, Li, Changjian
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866910891092475904
author Wei, Guangshun
Feng, Yuan
Ma, Long
Wang, Chen
Zhou, Yuanfeng
Li, Changjian
author_facet Wei, Guangshun
Feng, Yuan
Ma, Long
Wang, Chen
Zhou, Yuanfeng
Li, Changjian
contents This paper presents PCDreamer, a novel method for point cloud completion. Traditional methods typically extract features from partial point clouds to predict missing regions, but the large solution space often leads to unsatisfactory results. More recent approaches have started to use images as extra guidance, effectively improving performance, but obtaining paired data of images and partial point clouds is challenging in practice. To overcome these limitations, we harness the relatively view-consistent multi-view diffusion priors within large models, to generate novel views of the desired shape. The resulting image set encodes both global and local shape cues, which are especially beneficial for shape completion. To fully exploit the priors, we have designed a shape fusion module for producing an initial complete shape from multi-modality input (i.e.,, images and point clouds), and a follow-up shape consolidation module to obtain the final complete shape by discarding unreliable points introduced by the inconsistency from diffusion priors. Extensive experimental results demonstrate our superior performance, especially in recovering fine details.
format Preprint
id arxiv_https___arxiv_org_abs_2411_19036
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle PCDreamer: Point Cloud Completion Through Multi-view Diffusion Priors
Wei, Guangshun
Feng, Yuan
Ma, Long
Wang, Chen
Zhou, Yuanfeng
Li, Changjian
Computer Vision and Pattern Recognition
Graphics
This paper presents PCDreamer, a novel method for point cloud completion. Traditional methods typically extract features from partial point clouds to predict missing regions, but the large solution space often leads to unsatisfactory results. More recent approaches have started to use images as extra guidance, effectively improving performance, but obtaining paired data of images and partial point clouds is challenging in practice. To overcome these limitations, we harness the relatively view-consistent multi-view diffusion priors within large models, to generate novel views of the desired shape. The resulting image set encodes both global and local shape cues, which are especially beneficial for shape completion. To fully exploit the priors, we have designed a shape fusion module for producing an initial complete shape from multi-modality input (i.e.,, images and point clouds), and a follow-up shape consolidation module to obtain the final complete shape by discarding unreliable points introduced by the inconsistency from diffusion priors. Extensive experimental results demonstrate our superior performance, especially in recovering fine details.
title PCDreamer: Point Cloud Completion Through Multi-view Diffusion Priors
topic Computer Vision and Pattern Recognition
Graphics
url https://arxiv.org/abs/2411.19036