3D Object Detection with Attention:Shell-Based Modeling

作     者：Xiaorui Zhang Ziquan Zhao Wei Sun Qi Cui 

作者机构：School of Computer and SoftwareNanjing University of Information Science&TechnologyNanjing210044China Wuxi Research InstituteNanjing University of Information Science&TechnologyWuxi214100China Engineering Research Center of Digital ForensicsMinistry of EducationJiangsu Engineering Center of Network MonitoringNanjing210044China Jiangsu Collaborative Innovation Center of Atmospheric Environment and Equipment Technology(CICAEET)Nanjing University of Information Science&TechnologyNanjing210044China School of AutomationNanjing University of Information Science&TechnologyNanjing210044China Department of Electrical and Computer EngineeringUniversity of WindsorWindsorN9B 3P4Canada 

出 版 物：《Computer Systems Science & Engineering》 (计算机系统科学与工程（英文）)

年 卷 期：2023年第46卷第7期

页      面：537-550页

核心收录：

学科分类：0810[工学-信息与通信工程] 08[工学] 0804[工学-仪器科学与技术] 0811[工学-控制科学与工程] 0812[工学-计算机科学与技术（可授工学、理学学位）] 

基　　金：This work was supported,in part,by the National Nature Science Foundation of China under grant numbers 62272236 in part,by the Natural Science Foundation of Jiangsu Province under grant numbers BK20201136,BK20191401 in part,by the Priority Academic Program Development of Jiangsu Higher Education Institutions(PAPD)fund. 

主　　题：3D object detection autonomous driving point cloud shell-based modeling self-attention mechanism 

摘      要：LIDAR point cloud-based 3D object detection aims to sense the surrounding environment by anchoring objects with the Bounding Box(BBox).However,under the three-dimensional space of autonomous driving scenes,the previous object detection methods,due to the pre-processing of the original LIDAR point cloud into voxels or pillars,lose the coordinate information of the original point cloud,slow detection speed,and gain inaccurate bounding box positioning.To address the issues above,this study proposes a new two-stage network structure to extract point cloud features directly by PointNet++,which effectively preserves the original point cloud coordinate information.To improve the detection accuracy,a shell-based modeling method is proposed.It roughly determines which spherical shell the coordinates belong to.Then,the results are refined to ground truth,thereby narrowing the localization range and improving the detection accuracy.To improve the recall of 3D object detection with bounding boxes,this paper designs a self-attention module for 3D object detection with a skip connection structure.Some of these features are highlighted by weighting them on the feature dimensions.After training,it makes the feature weights that are favorable for object detection get larger.Thus,the extracted features are more adapted to the object detection task.Extensive comparison experiments and ablation experiments conducted on the KITTI dataset verify the effectiveness of our proposed method in improving recall and precision.