青岛大学学报（自然科学版）2025，Vol.38Issue(3)：28-34,7.DOI:10.3969/j.issn.1006-1037.2025.03.05
基于自注意力机制的区域提案优化3D目标检测网络
3D Object Detection Network Based on Self-attention Mechanism for Regional Proposal Optimization
摘要
Abstract
In the proposal refinement stage of the object detection algorithms there were two key problems including insufficient capture of contextual information and inadequate modeling of feature correlations.The improvement of detection accuracy was restricted by these problems.To address these challengs,a 3D object detection network for region pro-posal optimization,namely Proposal Refinement Optimization-RCNN(PRO-RCNN),based on the self-attention mechanism was proposed.On the basis of the PV-RCNN is model,the Transformer model was utilized to dynamically learn feature weights,capture rich contextual information,and model the correlations between objects,thereby optimi-zing the proposal generation results.The experimental results on the KITTI dataset show that the accuracy of PRO-RCNN is improved in all test difficulty categories,the average precision of pedestrian category detection is increased by 2%to 3%.关键词
目标检测/深度学习/点云/自注意力/神经网络Key words
object detection/deep learning/point cloud/self-attention/neural network分类
信息技术与安全科学引用本文复制引用
张欣,毕博学,昝国宽,赵俊莉,万志波..基于自注意力机制的区域提案优化3D目标检测网络[J].青岛大学学报（自然科学版）,2025,38(3):28-34,7.基金项目
国家自然科学基金(批准号:62172247)资助. （批准号:62172247）
