Point2Node: Correlation Learning of Dynamic-Node for Point Cloud Feature Modeling

Han, Wenkai; Wen, Chenglu; Wang, Cheng; Li, Xin; Li, Qing

doi:10.1609/aaai.v34i07.6725

Cited by 64 publications

(42 citation statements)

References 21 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…By default, graph-based methods capture the correlation between the points using edges. Point2Node [81] exploits not only the local correlation but also the non-local correlation between points and has better performance in terms semantic segmentation. The performance of the methods discussed in classification, segmentation, and object detection applications are shown in Section 7.…”

Section: Discussionmentioning

confidence: 99%

“…Graph based approaches were proposed in [78][79][80][81]; others include [82][83][84][85]. Graph-based approaches represent point clouds with a graph structure by treating each point as a node.…”

Section: Graph-based Approachesmentioning

confidence: 99%

“…The kd-network captures hierarchical representations along the depth of the kd-tree; however, because of the tree design, nodes at the same depth level do not capture overlapping receptive fields. In [79][80][81], the authors use a method based on the typical graph network G = {V, E} whose vertices V represent the points, and the edges E are represented as a V × V matrix. In DGCNN [79], edge convolution is proposed.…”

Section: Graph-based Approachesmentioning

confidence: 99%

“…In Point2Node [81], the authors proposed a graph network that fully explores not only the local correlation but also non-local correlation. The correlation is explored in three ways: self-correlation, which explores the channel-wise correlation of a node's feature; local correlation, which explores the local dependency among nodes in a local region; and non-local correlation, which is used to capture better global features by considering long-range local features.…”

Section: Graph-based Approachesmentioning

confidence: 99%

See 3 more Smart Citations

Review: Deep Learning on 3D Point Clouds

et al. 2020

Self Cite

View full text Add to dashboard Cite

A point cloud is a set of points defined in a 3D metric space. Point clouds have become one of the most significant data formats for 3D representation and are gaining increased popularity as a result of the increased availability of acquisition devices, as well as seeing increased application in areas such as robotics, autonomous driving, and augmented and virtual reality. Deep learning is now the most powerful tool for data processing in computer vision and is becoming the most preferred technique for tasks such as classification, segmentation, and detection. While deep learning techniques are mainly applied to data with a structured grid, the point cloud, on the other hand, is unstructured. The unstructuredness of point clouds makes the use of deep learning for its direct processing very challenging. This paper contains a review of the recent state-of-the-art deep learning techniques, mainly focusing on raw point cloud data. The initial work on deep learning directly with raw point cloud data did not model local regions; therefore, subsequent approaches model local regions through sampling and grouping. More recently, several approaches have been proposed that not only model the local regions but also explore the correlation between points in the local regions. From the survey, we conclude that approaches that model local regions and take into account the correlation between points in the local regions perform better. Contrary to existing reviews, this paper provides a general structure for learning with raw point clouds, and various methods were compared based on the general structure. This work also introduces the popular 3D point cloud benchmark datasets and discusses the application of deep learning in popular 3D vision tasks, including classification, segmentation, and detection.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Graph-based Approachesmentioning

confidence: 99%

Section: Graph-based Approachesmentioning

confidence: 99%

Section: Graph-based Approachesmentioning

confidence: 99%

See 2 more Smart Citations

Review: Deep Learning on 3D Point Clouds

et al. 2020

Self Cite

View full text Add to dashboard Cite

show abstract

“…RandLA-Net [20] proposes a local feature enhancement module, which extracts local geometric structures by increasing the visible area of each point. As an end-to-end graph network, Point2Node [21] learns the relationship between points from different scales and strengthens geometric connections in high-dimensional space with a self-adaptive enhanced feature mechanism.…”

Section: Introductionmentioning

confidence: 99%

Local Feature Extraction Network for Point Cloud Analysis

Zhou¹,

Tai²,

Chen³

et al. 2021

Symmetry

View full text Add to dashboard Cite

Geometric feature extraction of 3D point clouds plays an important role in many 3D computer vision applications such as region labeling, 3D reconstruction, object segmentation, and recognition. However, hand-designed features on point clouds lack semantic information, so cannot meet these requirements. In this paper, we propose local feature extraction network (LFE-Net) which focus on extracting local feature for point clouds analysis. Such geometric features learning from a relation of local points can be used in a variety of shape analysis problems such as classification, part segmentation, and point matching. LFE-Net consists of local geometric relation (LGR) module which aims to learn a high-dimensional local feature to express the relation between points and their neighbors. Benefiting from the additional singular values of local points and hierarchical neural networks, the learned local features are robust to permutation and rigid transformation so that they can be transformed into 3D descriptors. Moreover, we embed prior spatial information of the local points into the sub-features for combining features from multiple levels. LFE-Net achieves state-of-the-art performances on standard benchmarks including ModelNet40, ShapeNetPart.

show abstract

SPNet: Multi-shell Kernel Convolution for Point Cloud Semantic Segmentation

Fan²,

Wang³

et al. 2021

Advances in Visual Computing

View full text Add to dashboard Cite

Point2Node: Correlation Learning of Dynamic-Node for Point Cloud Feature Modeling

Cited by 64 publications

References 21 publications

Review: Deep Learning on 3D Point Clouds

Review: Deep Learning on 3D Point Clouds

Local Feature Extraction Network for Point Cloud Analysis

SPNet: Multi-shell Kernel Convolution for Point Cloud Semantic Segmentation

Contact Info

Product

Resources

About