DGANet: A Dilated Graph Attention-Based Network for Local Feature Extraction on 3D Point Clouds
Feature extraction on point clouds is an essential task when analyzing and processing point clouds of 3D scenes. However, there still remains a challenge to adequately exploit local fine-grained features on point cloud data due to its irregular and unordered structure in a 3D space. To alleviate thi...
Main Authors: | , , , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
MDPI AG
2021-09-01
|
Series: | Remote Sensing |
Subjects: | |
Online Access: | https://www.mdpi.com/2072-4292/13/17/3484 |
_version_ | 1797520927912624128 |
---|---|
author | Jie Wan Zhong Xie Yongyang Xu Ziyin Zeng Ding Yuan Qinjun Qiu |
author_facet | Jie Wan Zhong Xie Yongyang Xu Ziyin Zeng Ding Yuan Qinjun Qiu |
author_sort | Jie Wan |
collection | DOAJ |
description | Feature extraction on point clouds is an essential task when analyzing and processing point clouds of 3D scenes. However, there still remains a challenge to adequately exploit local fine-grained features on point cloud data due to its irregular and unordered structure in a 3D space. To alleviate this problem, a Dilated Graph Attention-based Network (DGANet) with a certain feature for learning ability is proposed. Specifically, we first build a local dilated graph-like region for each input point to establish the long-range spatial correlation towards its corresponding neighbors, which allows the proposed network to access a wider range of geometric information of local points with their long-range dependencies. Moreover, by integrating the dilated graph attention module (DGAM) implemented by a novel offset–attention mechanism, the proposed network promises to highlight the differing importance on each edge of the constructed local graph to uniquely learn the discrepancy feature of geometric attributes between the connected point pairs. Finally, all the learned edge attention features are further aggregated, allowing the most significant geometric feature representation of local regions by the graph–attention pooling to fully extract local detailed features for each point. The validation experiments using two challenging benchmark datasets demonstrate the effectiveness and powerful generation ability of our proposed DGANet in both 3D object classification and segmentation tasks. |
first_indexed | 2024-03-10T08:05:26Z |
format | Article |
id | doaj.art-e929d26aba7a4a85958769910946a7f2 |
institution | Directory Open Access Journal |
issn | 2072-4292 |
language | English |
last_indexed | 2024-03-10T08:05:26Z |
publishDate | 2021-09-01 |
publisher | MDPI AG |
record_format | Article |
series | Remote Sensing |
spelling | doaj.art-e929d26aba7a4a85958769910946a7f22023-11-22T11:09:37ZengMDPI AGRemote Sensing2072-42922021-09-011317348410.3390/rs13173484DGANet: A Dilated Graph Attention-Based Network for Local Feature Extraction on 3D Point CloudsJie Wan0Zhong Xie1Yongyang Xu2Ziyin Zeng3Ding Yuan4Qinjun Qiu5Key Laboratory of Geological and Evaluation of Ministry of Education, China University of Geosciences, Wuhan 430074, ChinaDepartment of Geography and Information Engineering, China University of Geosciences, Wuhan 430074, ChinaDepartment of Geography and Information Engineering, China University of Geosciences, Wuhan 430074, ChinaDepartment of Geography and Information Engineering, China University of Geosciences, Wuhan 430074, ChinaDepartment of Geography and Information Engineering, China University of Geosciences, Wuhan 430074, ChinaNational Engineering Research Center of Geographic Information System, Wuhan 430074, ChinaFeature extraction on point clouds is an essential task when analyzing and processing point clouds of 3D scenes. However, there still remains a challenge to adequately exploit local fine-grained features on point cloud data due to its irregular and unordered structure in a 3D space. To alleviate this problem, a Dilated Graph Attention-based Network (DGANet) with a certain feature for learning ability is proposed. Specifically, we first build a local dilated graph-like region for each input point to establish the long-range spatial correlation towards its corresponding neighbors, which allows the proposed network to access a wider range of geometric information of local points with their long-range dependencies. Moreover, by integrating the dilated graph attention module (DGAM) implemented by a novel offset–attention mechanism, the proposed network promises to highlight the differing importance on each edge of the constructed local graph to uniquely learn the discrepancy feature of geometric attributes between the connected point pairs. Finally, all the learned edge attention features are further aggregated, allowing the most significant geometric feature representation of local regions by the graph–attention pooling to fully extract local detailed features for each point. The validation experiments using two challenging benchmark datasets demonstrate the effectiveness and powerful generation ability of our proposed DGANet in both 3D object classification and segmentation tasks.https://www.mdpi.com/2072-4292/13/17/34843D point cloudslocal feature extractiondeep learninggraph attention mechanism |
spellingShingle | Jie Wan Zhong Xie Yongyang Xu Ziyin Zeng Ding Yuan Qinjun Qiu DGANet: A Dilated Graph Attention-Based Network for Local Feature Extraction on 3D Point Clouds Remote Sensing 3D point clouds local feature extraction deep learning graph attention mechanism |
title | DGANet: A Dilated Graph Attention-Based Network for Local Feature Extraction on 3D Point Clouds |
title_full | DGANet: A Dilated Graph Attention-Based Network for Local Feature Extraction on 3D Point Clouds |
title_fullStr | DGANet: A Dilated Graph Attention-Based Network for Local Feature Extraction on 3D Point Clouds |
title_full_unstemmed | DGANet: A Dilated Graph Attention-Based Network for Local Feature Extraction on 3D Point Clouds |
title_short | DGANet: A Dilated Graph Attention-Based Network for Local Feature Extraction on 3D Point Clouds |
title_sort | dganet a dilated graph attention based network for local feature extraction on 3d point clouds |
topic | 3D point clouds local feature extraction deep learning graph attention mechanism |
url | https://www.mdpi.com/2072-4292/13/17/3484 |
work_keys_str_mv | AT jiewan dganetadilatedgraphattentionbasednetworkforlocalfeatureextractionon3dpointclouds AT zhongxie dganetadilatedgraphattentionbasednetworkforlocalfeatureextractionon3dpointclouds AT yongyangxu dganetadilatedgraphattentionbasednetworkforlocalfeatureextractionon3dpointclouds AT ziyinzeng dganetadilatedgraphattentionbasednetworkforlocalfeatureextractionon3dpointclouds AT dingyuan dganetadilatedgraphattentionbasednetworkforlocalfeatureextractionon3dpointclouds AT qinjunqiu dganetadilatedgraphattentionbasednetworkforlocalfeatureextractionon3dpointclouds |