GA-RCNN:Graph self-attention feature extraction for 3D object detection

Author:

Yi Yangyang1,Yu Long1,Tian Shengwei1,Gao Xuezhuang1,Li Jie1,Zhao Xingang2

Affiliation:

1. Collage of Software, Xinjiang University, Xinjiang Uyghur Autonomous Region, China

2. Land and Property Department, China Railway Urumqi Bureau Group Co., Xinjiang Uyghur Autonomous Region, China

Abstract

In recent years, 3D object detection based on LiDAR point clouds is a key component of autonomous driving. In pursuit of enhancing the accuracy of 3D point cloud feature extraction and point cloud detection, this paper introduces a novel 3D object detection model, termed as Graph Self-Attention-RCNN (GA-RCNN). This model is designed to integrate voxel information and point location information, enhancing the quality of 3D object proposals while maintaining contextual accuracy. The first stage rectifies the previous approach that relied on local features for preselected boxes, overlooking crucial global contextual information. An improved method is suggested in this work, utilizing BEV to capture long-range dependencies via a cross-attention mechanism. The second stage addresses the overreliance on local neighborhood point feature extraction. The Graph Self-Attention Pooling method is proposed, characterized by its dynamic computation of contribution weights for inputs. This enhances the model’s flexibility and generalization performance. Extensive evaluations on KITTI and Waymo datasets demonstrate GA-RCNN’s superior accuracy compared to other methods, affirming its efficacy in 3D object detection.

Publisher

IOS Press

Subject

Artificial Intelligence,General Engineering,Statistics and Probability

Reference14 articles.

1. Drone detection using sparse lidarmeasurements;Dogru;IEEE Robotics and Automation Letters,2022

2. Efficientmspso sampling for object detection and 6-d pose estimation in 3-dscenes;Xing;IEEE Transactions on Industrial Electronics,2021

3. Second: Sparsely embedded convolutionaldetection;Yan;Sensors,2018

4. Pointnet++: Deep hierarchicalfeature learning on point sets in a metric space;Qi;Advances inNeural Information Processing Systems

5. Attention is all you need;Vaswani;Advances in Neural Information Processing Systems

同舟云学术

1.学者识别学者识别

2.学术分析学术分析

3.人才评估人才评估

"同舟云学术"是以全球学者为主线,采集、加工和组织学术论文而形成的新型学术文献查询和分析系统,可以对全球学者进行文献检索和人才价值评估。用户可以通过关注某些学科领域的顶尖人物而持续追踪该领域的学科进展和研究前沿。经过近期的数据扩容,当前同舟云学术共收录了国内外主流学术期刊6万余种,收集的期刊论文及会议论文总量共计约1.5亿篇,并以每天添加12000余篇中外论文的速度递增。我们也可以为用户提供个性化、定制化的学者数据。欢迎来电咨询!咨询电话:010-8811{复制后删除}0370

www.globalauthorid.com

TOP

Copyright © 2019-2024 北京同舟云网络信息技术有限公司
京公网安备11010802033243号  京ICP备18003416号-3