|本期目录/Table of Contents|

[1]葛旭阳,黄尚锋,蔡国榕.基于三维立方体损失的点云目标检测算法[J].集美大学学报(自然科学版),2021,26(3):280-288.
　GE Xuyang,HUANG Shangfeng,CAI Guorong.Point Cloudobject Detection Algorithm Based on 3D Cube Loss[J].Journal of Jimei University,2021,26(3):280-288.
点击复制

基于三维立方体损失的点云目标检测算法(PDF)
分享到：

《集美大学学报（自然科学版）》[ISSN:1007-7405/CN:35-1186/N]

卷:
第26卷
期数:
2021年第3期
页码:
280-288
栏目:
数理科学与信息工程
出版日期:
2021-05-28

文章信息/Info

Title:
Point Cloudobject Detection Algorithm Based on 3D Cube Loss
作者:
葛旭阳1; 黄尚锋2; 蔡国榕2
(1.集美大学理学院,福建 厦门 361021;2.集美大学计算机工程学院,福建 厦门 361021)
Author(s):
GE Xuyang1; HUANG Shangfeng2; CAI Guorong2
(1.School of Science,Jimei University,Xiamen 361021,China；2.College of Computer Engineering，Jimei University，Xiamen 361021，China)
关键词:
目标检测; IoU; 3D_CGIoU; 度量; 损失
Keywords:
object detection; IoU; 3D_CGIoU; metric; loss
分类号:
-
DOI:
-
文献标志码:
-
摘要:
常用的优化回归边界框参数的损失并不等价于最大化IoU指标，并且IoU作为回归损失，在边界框不重叠的情况下进行优化是不可行的。为了解决这个问题，在IoU基础上加入外接框的计算，将所得结果(3D_CGIoU)作为损失纳入到目前主流的三维目标检测框架中，并在数据集KITTI和ScanNet上进行实验。实验结果显示检测的平均精度得到了提升，表明该方法是有效的。
Abstract:
Optimizing the commonly used losses for regressing the parameters of a bounding box is not equivalent to maximizing the IoU index.And IoU as a regression loss,it is not feasible to optimize when boundary boxes do not overlap.In order to solve the above problems,the calculation of the circumscribed box is added to the basis of IoU(3D_CGIoU) as a new metric and new loss.Incorporating 3D_CGIoU as a loss into the stateoftheart 3D object detection framework,and experiments were conducted on the KITTI dataset and the ScanNet dataset respectively.The improvement of the average accuracy (AP) of the detection shows the effectiveness of the method.

参考文献/References:

-

相似文献/References:

[1]郑东强,周海峰,曾伟民,等.基于多维空间离散基准的目标检测网络设计[J].集美大学学报(自然科学版),2024,29(3):282.
　ZHENG Dongqiang,ZHOU Haifeng,ZENG Weimin,et al.The Design of Object Detection Network Based on Multidimensional Space Discrete Datum[J].Journal of Jimei University,2024,29(3):282.

备注/Memo

备注/Memo:
-
更新日期/Last Update: 2021-07-13