A Strong View-Free Baseline Approach for Single-View Image Guided Point Cloud Completion

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Lin, Fangzhou, Dai, Zilin, Sanku, Rigved, Hou, Songlin, Yamada, Kazunori D, Zhang, Haichong K., Zhang, Ziming
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866909944966545408
author Lin, Fangzhou
Dai, Zilin
Sanku, Rigved
Hou, Songlin
Yamada, Kazunori D
Zhang, Haichong K.
Zhang, Ziming
author_facet Lin, Fangzhou
Dai, Zilin
Sanku, Rigved
Hou, Songlin
Yamada, Kazunori D
Zhang, Haichong K.
Zhang, Ziming
contents The single-view image guided point cloud completion (SVIPC) task aims to reconstruct a complete point cloud from a partial input with the help of a single-view image. While previous works have demonstrated the effectiveness of this multimodal approach, the fundamental necessity of image guidance remains largely unexamined. To explore this, we propose a strong baseline approach for SVIPC based on an attention-based multi-branch encoder-decoder network that only takes partial point clouds as input, view-free. Our hierarchical self-fusion mechanism, driven by cross-attention and self-attention layers, effectively integrates information across multiple streams, enriching feature representations and strengthening the networks ability to capture geometric structures. Extensive experiments and ablation studies on the ShapeNet-ViPC dataset demonstrate that our view-free framework performs superiorly to state-of-the-art SVIPC methods. We hope our findings provide new insights into the development of multimodal learning in SVIPC. Our demo code will be available at https://github.com/Zhang-VISLab.
format Preprint
id arxiv_https___arxiv_org_abs_2506_15747
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle A Strong View-Free Baseline Approach for Single-View Image Guided Point Cloud Completion
Lin, Fangzhou
Dai, Zilin
Sanku, Rigved
Hou, Songlin
Yamada, Kazunori D
Zhang, Haichong K.
Zhang, Ziming
Computer Vision and Pattern Recognition
Image and Video Processing
The single-view image guided point cloud completion (SVIPC) task aims to reconstruct a complete point cloud from a partial input with the help of a single-view image. While previous works have demonstrated the effectiveness of this multimodal approach, the fundamental necessity of image guidance remains largely unexamined. To explore this, we propose a strong baseline approach for SVIPC based on an attention-based multi-branch encoder-decoder network that only takes partial point clouds as input, view-free. Our hierarchical self-fusion mechanism, driven by cross-attention and self-attention layers, effectively integrates information across multiple streams, enriching feature representations and strengthening the networks ability to capture geometric structures. Extensive experiments and ablation studies on the ShapeNet-ViPC dataset demonstrate that our view-free framework performs superiorly to state-of-the-art SVIPC methods. We hope our findings provide new insights into the development of multimodal learning in SVIPC. Our demo code will be available at https://github.com/Zhang-VISLab.
title A Strong View-Free Baseline Approach for Single-View Image Guided Point Cloud Completion
topic Computer Vision and Pattern Recognition
Image and Video Processing
url https://arxiv.org/abs/2506.15747