APSIPA Transactions on Signal and Information Processing > Vol 14 > Issue 2

JointFormer: Joint-Enhanced 3D Human Point Cloud Completion Based on Transformer

Min Zhou, Shanghai University, China, Jieyu Chen, Shanghai University, China, Xinpeng Huang, Shanghai University, China, Ping An, Shanghai University, China, anping@shu.edu.cn
 
Suggested Citation
Min Zhou, Jieyu Chen, Xinpeng Huang and Ping An (2025), "JointFormer: Joint-Enhanced 3D Human Point Cloud Completion Based on Transformer", APSIPA Transactions on Signal and Information Processing: Vol. 14: No. 2, e104. http://dx.doi.org/10.1561/116.20240041

Publication Date: 23 Apr 2025
© 2025 M. Zhou, J. Chen, X. Huang and P. An
 
Subjects
Deep learning,  3D reconstruction and image-based modeling,  Sensor and multiple source signal processing
 
Keywords
Point cloud completionhuman point cloudtransformerjoints estimationspatial attention
 

Share

Open Access

This is published under the terms of CC BY-NC.

Downloaded: 97 times

In this article:
Introduction 
Related Work 
Proposed Method 
Experiments 
Conclusions 
References 

Abstract

Human point cloud completion is a challenging yet indispensable task, devoted to filling missing parts in the collected incomplete point clouds. Existing methods overly rely on features extracted from surface points, neglecting the intrinsic joints information point clouds possess. To address this problem, we propose a new network with an encoder-decoder framework, named JointFormer. Firstly, we design a joint-enhanced encoder that provides more prior guidance on the overall structure of the partial input. Then, a generator is employed to generate sparse but complete point clouds. Finally, a decoder refines the rough point clouds into complete and dense human body point clouds in a coarse-to-fine manner. Moreover, combining transformer with the Convolutional Block Attention Module (CBAM), we design the Channel-Spatial Attention Transformer (CSAT) to better capture point cloud spatial relationships. Quantitative and qualitative evaluations demonstrate that JointFormer outperforms the state-of-the-art completion method on our two human body point cloud datasets.

DOI:10.1561/116.20240041

Companion

APSIPA Transactions on Signal and Information Processing Special Issue - Three-dimensional Point Cloud Data Modeling, Processing, and Analysis
See the other articles that are part of this special issue.