液晶与显示2024，Vol.39Issue(1)：79-88,10.DOI:10.37188/CJLCD.2023-0058
基于知识蒸馏和定位引导的Pointpillars点云检测网络
Pointpillars point cloud detection network based on knowledge distillation and location guidance
摘要
Abstract
Lidar data is widely used in 3D target detection tasks due to its geometric characteristics.Due to the sparsity and irregularity of point cloud data,it is difficult to achieve the balance between the quality of feature extraction and the speed of reasoning.In this paper,a three-dimensional target detection algorithm based on body-column feature coding is proposed.Based on Pointpillars network,the Teacher-Student model framework is designed to distill the regression frame scale,increase distillation loss,optimize the training network model,and improve the quality of feature extraction.In order to further improve the model detection effect,the positioning guidance classification item is designed to increase the correlation between classification prediction and regression prediction,and improve the object recognition accuracy.The improvement of this network does not introduce additional network embedding.The experimental results of the algorithm on the KITTI dataset show that the average accuracy of the reference network in 3D mode is improved from 60.65%to 64.69%,and the average accuracy of the aerial view mode is improved from 67.74%to 70.24%.The model reasoning speed is 45 FPS,which meets the real-time requirements while improving the detection accuracy.关键词
激光点云/三维目标检测/知识蒸馏/分类置信度Key words
laser point cloud/3D object detection/knowledge distillation/classification confidence分类
信息技术与安全科学引用本文复制引用
赵晶,李少博,郭杰龙,俞辉,张剑锋,李杰..基于知识蒸馏和定位引导的Pointpillars点云检测网络[J].液晶与显示,2024,39(1):79-88,10.基金项目
福建省科技计划(No.2021T3003) （No.2021T3003）
泉州市科技计划(No.2021C065L) （No.2021C065L）
福建省科技厅自然科学基金(No.2020J01285,No.2022J05285)Supported by Fujian Provincial Science and Technology Plan(No.2021T3003) （No.2020J01285,No.2022J05285）
Quanzhou Science and Technology Plan(No.2021C065L) （No.2021C065L）
Natural Science Foundation of Fujian Provincial Department of Science and Technology(No.2020J01285,No.2022J05285) （No.2020J01285,No.2022J05285）
