Abstract:Three-dimensional (3D) point clouds have a wide range of applications in the field of 3D vision. The quality of the acquired point cloud data considerably impacts the subsequent work of point cloud processing. Due to the sparsity and irregularity of point cloud data, processing point cloud data has always been challenging. However, existing deep learning-based point cloud dense reconstruction methods suffer from excessive smoothing of reconstruction results and too many outliers. The reason for this is that it is not possible to extract features for local and global features at different scales and provide different levels of attention to different regions in order to obtain long-distance dependence for dense reconstruction. In this paper, we use a parallel multi-scale feature extraction module based on graph convolution and an upsampling method with an added multi-head attention mechanism to process sparse and irregular point cloud data to obtain extended point clouds. Specifically, a point cloud training patch with 256 points is inputted. The PMS module uses three residual connections in the multi-scale feature extraction stage. Each PMS module consists of three parallel DenseGCN modules with different size convolution kernels and different averaging pooling sizes. The local and global feature information of the augmented receptive field is extracted efficiently. The scale information is obtained by averaging the different pooled augmented receptive fields. The scale information was obtained using the different average pooled augmented receptive fields. The upsampling stage uses an upsampling rate of r=4, The self-attentive features with a different focus on different point cloud data regions obtained by fusing different weights make the feature representation more diverse. This operation avoids the bias of one attention, and each focuses on extracting valuable fine-grained feature information. Finally, the coordinate reconstruction module obtains 1024 dense point cloud data. Experiments show that the proposed method demonstrates good evaluation metrics and performance and is able to obtain better visual quality. The problems of over-smoothing and excessive outliers are effectively mitigated, and the obtained sparse point cloud is more dense.

Multi-space and detail-supplemented attention network for point cloud completion

Dual-scale Point Cloud Completion Network Based on High-Frequency Feature Fusion

Point Cloud Completion Cascade Optimization Network Based on Feature Fusion

Adaptive Recurrent Forward Network for Dense Point Cloud Completion

Point Cloud Completion via Multi-Scale Edge Convolution and Attention

Aggregating Feature Point Cloud for Depth Completion

Research on Multi-modal Point Cloud Completion Task

PointMS: Semantic Segmentation for Point Cloud Based on Multi-scale Directional Convolution

A coarse-to-fine point completion network with details compensation and structure enhancement

Fast point completion network

N-DPC: Dense 3D Point Cloud Completion Based on Improved Multi-Stage Network

End-to-End Point Cloud Completion Network with Attention Mechanism

Multi-feature fusion point cloud completion network

Multistage Adaptive Point-Growth Network for Dense Point Cloud Completion

Point Cloud Completion Based on Nonlocal Neural Networks with Adaptive Sampling

Completing Partial Point Clouds with Outliers by Collaborative Completion and Segmentation

PointAttN: You Only Need Attention for Point Cloud Completion

Sparse 3D Point Cloud Parallel Multi-Scale Feature Extraction and Dense Reconstruction with Multi-Headed Attentional Upsampling

3DPCTN: Two 3D Local-Object Point-Cloud-Completion Transformer Networks Based on Self-Attention and Multi-Resolution

2D-3DMatchingNet: Multimodal Point Completion with 2D Geometry Matching

Point Cloud Completion by Skip-attention Network with Hierarchical Folding