Multi-feature enhancement based on sparse networks for single-stage 3D object detection
In the field of autonomous driving, the accuracy and real-time requirements for 3D object detection technology continue to improve, which is directly related to the commercialization process and market popularity of autonomous vehicles. Despite the efficiency of pillar-based coding for onboard syste...
Saved in:
| Main Authors: | , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Elsevier
2025-01-01
|
| Series: | Alexandria Engineering Journal |
| Subjects: | |
| Online Access: | http://www.sciencedirect.com/science/article/pii/S111001682401216X |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1841524923098464256 |
|---|---|
| author | Zunwang Ke Chenyu Lin Tao Zhang Tingting Jia Minghua Du Gang Wang Yugui Zhang |
| author_facet | Zunwang Ke Chenyu Lin Tao Zhang Tingting Jia Minghua Du Gang Wang Yugui Zhang |
| author_sort | Zunwang Ke |
| collection | DOAJ |
| description | In the field of autonomous driving, the accuracy and real-time requirements for 3D object detection technology continue to improve, which is directly related to the commercialization process and market popularity of autonomous vehicles. Despite the efficiency of pillar-based coding for onboard systems, it falls short in terms of accuracy and the reduction of incorrect positives. In this paper, we will examine how to solve the problem of high incorrect rate and low accuracy of existing methods. Firstly, a MAP coding module is introduced to optimize previous point cloud feature coding modules, allowing for the efficient extraction of fine-grained features from point cloud data. Then, we introduce an innovative sparse dual attention (SDA) to efficiently filter out irrelevant details in feature extraction, thereby improving the pertinence and efficiency of information extraction. Finally, to address the potential loss of information from single local feature extraction, a local and global fusion module (CTGC) is introduced. Our method has proactively demonstrated its efficiency and accuracy through rigorous experimentation across diverse datasets. Analysis of the results leads to the conclusion that our solutions provide accurate and robust detection results. Code will be available at https://github.com/lcy199905/MyOpenPCDet.git. |
| format | Article |
| id | doaj-art-b6e4d613179449f592b81df1d32ee29e |
| institution | Kabale University |
| issn | 1110-0168 |
| language | English |
| publishDate | 2025-01-01 |
| publisher | Elsevier |
| record_format | Article |
| series | Alexandria Engineering Journal |
| spelling | doaj-art-b6e4d613179449f592b81df1d32ee29e2025-01-18T05:03:39ZengElsevierAlexandria Engineering Journal1110-01682025-01-01111123135Multi-feature enhancement based on sparse networks for single-stage 3D object detectionZunwang Ke0Chenyu Lin1Tao Zhang2Tingting Jia3Minghua Du4Gang Wang5Yugui Zhang6School of Software, Xinjiang University, Urumqi 830091, ChinaSchool of Software, Xinjiang University, Urumqi 830091, ChinaSchool of Software, Xinjiang University, Urumqi 830091, ChinaDepartment of Stomatology, The First Medical Centre, Chinese PLA General Hospital, Beijing 100853, ChinaDepartment of Stomatology, The First Medical Centre, Chinese PLA General Hospital, Beijing 100853, ChinaSchool of Computing and Data Engineering, NingboTech University, Ningbo 315100, ChinaInstitute of Semiconductors, Chinese Academy of Sciences, Beijing 100083, China; Corresponding author.In the field of autonomous driving, the accuracy and real-time requirements for 3D object detection technology continue to improve, which is directly related to the commercialization process and market popularity of autonomous vehicles. Despite the efficiency of pillar-based coding for onboard systems, it falls short in terms of accuracy and the reduction of incorrect positives. In this paper, we will examine how to solve the problem of high incorrect rate and low accuracy of existing methods. Firstly, a MAP coding module is introduced to optimize previous point cloud feature coding modules, allowing for the efficient extraction of fine-grained features from point cloud data. Then, we introduce an innovative sparse dual attention (SDA) to efficiently filter out irrelevant details in feature extraction, thereby improving the pertinence and efficiency of information extraction. Finally, to address the potential loss of information from single local feature extraction, a local and global fusion module (CTGC) is introduced. Our method has proactively demonstrated its efficiency and accuracy through rigorous experimentation across diverse datasets. Analysis of the results leads to the conclusion that our solutions provide accurate and robust detection results. Code will be available at https://github.com/lcy199905/MyOpenPCDet.git.http://www.sciencedirect.com/science/article/pii/S111001682401216XPoint cloudObject detectionMAP codingSparse dual attentionLocal and global fusion |
| spellingShingle | Zunwang Ke Chenyu Lin Tao Zhang Tingting Jia Minghua Du Gang Wang Yugui Zhang Multi-feature enhancement based on sparse networks for single-stage 3D object detection Alexandria Engineering Journal Point cloud Object detection MAP coding Sparse dual attention Local and global fusion |
| title | Multi-feature enhancement based on sparse networks for single-stage 3D object detection |
| title_full | Multi-feature enhancement based on sparse networks for single-stage 3D object detection |
| title_fullStr | Multi-feature enhancement based on sparse networks for single-stage 3D object detection |
| title_full_unstemmed | Multi-feature enhancement based on sparse networks for single-stage 3D object detection |
| title_short | Multi-feature enhancement based on sparse networks for single-stage 3D object detection |
| title_sort | multi feature enhancement based on sparse networks for single stage 3d object detection |
| topic | Point cloud Object detection MAP coding Sparse dual attention Local and global fusion |
| url | http://www.sciencedirect.com/science/article/pii/S111001682401216X |
| work_keys_str_mv | AT zunwangke multifeatureenhancementbasedonsparsenetworksforsinglestage3dobjectdetection AT chenyulin multifeatureenhancementbasedonsparsenetworksforsinglestage3dobjectdetection AT taozhang multifeatureenhancementbasedonsparsenetworksforsinglestage3dobjectdetection AT tingtingjia multifeatureenhancementbasedonsparsenetworksforsinglestage3dobjectdetection AT minghuadu multifeatureenhancementbasedonsparsenetworksforsinglestage3dobjectdetection AT gangwang multifeatureenhancementbasedonsparsenetworksforsinglestage3dobjectdetection AT yuguizhang multifeatureenhancementbasedonsparsenetworksforsinglestage3dobjectdetection |