期刊文献+

图卷积神经网络在点云语义分割中的研究综述 被引量:4

Review of Graph Convolutional Neural Networks in Point Cloud Semantic Segmentation
下载PDF
导出
摘要 点云数据蕴含丰富的空间信息,可以通过激光雷达、3D传感器等设备大量采集,被广泛应用于自动驾驶、虚拟现实、城市规划和3D重建等领域。点云语义分割作为3D场景理解、识别和各种应用的基础而受到广泛关注。但不规则的点云数据无法直接作为传统卷积神经网络的输入,而图卷积神经网络可以利用图卷积算子直接对点云数据进行特征提取,使得图卷积神经网络已逐步成为点云语义分割领域的一个重要研究方向。基于此,对图卷积神经网络在3D点云语义分割应用中的研究进展进行综述,根据图卷积的类型对基于图卷积神经网络的点云语义分割方法进行分类,按照不同类别对比分析主流方法的模型架构及其特点,描述几个相关点云语义分割领域常用的公共数据集和评价指标,对点云语义分割方法进行总结和展望。 Point cloud contains rich spatial information,which can be collected in large quantities with lidar,3D sensors and other equipment,and is widely used in the fields of autonomous driving,virtual reality,urban planning and 3D reconstruction.Point cloud semantic segmentation has received wide attention as the basis for 3D scene understanding,recognition and various applications.However,irregular point cloud data cannot be directly used as the input of traditional convolutional neural networks,while graph convolutional neural networks can directly extract features from point cloud data using graph convolution operators,making graph convolutional neural networks have gradually become an important research direction in the field of point cloud semantic segmentation.Based on this,this paper reviews the research progress of graph convolutional neural network in 3D point cloud semantic segmentation application.Firstly,the point cloud semantic segmentation methods based on graph convolution are classified according to the types of graph convolution,and the model architectures and characteristics of mainstream methods are compared and analyzed according to different categories.Then,several common public datasets and evaluation metrics in the related point cloud semantic segmentation field are described.Finally,a summary and prospect of point cloud semantic segmentation methods are presented.
作者 张蕊 孟晓曼 曾志远 金玮 武益超 ZHANG Rui;MENG Xiaoman;ZENG Zhiyuan;JIN Wei;WU Yichao(School of Information Engineering,North China University of Water Resources and Electric Power,Zhengzhou 450046,China)
出处 《计算机工程与应用》 CSCD 北大核心 2022年第24期29-46,共18页 Computer Engineering and Applications
基金 2021年河南省科技攻关项目(202102210141)。
关键词 语义分割 3D点云 图卷积神经网络 深度学习 semantic segmentation 3D point cloud graph convolutional neural network deep learning
  • 相关文献

参考文献8

二级参考文献101

  • 1KRIZHEVSKY A,SUTSKEVER I,HINTON G E.Imagenet classification with deep convolutional neural networks[C]∥Advances in Neural Information Processing Systems.Red Hook,NY:Curran Associates,2012:1097-1105.
  • 2DAHL G E,YU D,DENG L,et al.Context-dependent pre-trained deep neural networks for large-vocabulary speech recognition[J].Audio,Speech,and Language Processing,IEEE Transactions on,2012,20(1):30-42.
  • 3ZEN H,SENIOR A,SCHUSTER M.Statistical parametric speech synthesis using deep neural networks[C]∥Acoustics,Speech and Signal Processing(ICASSP),20131EEE International Conference on.Piscataway,NJ:IEEE,2013:7962-7966.
  • 4BAHDANAU D,CHO K,BENGIO Y.Neural machine translation by jointly learning to align and translate[J].CoRR,2014:abs/1409.0473.
  • 5ZEILER M D,FERGUS R.Visualizing and understanding convolutional neural networks[J].CoRR,2013:abs/1311.2901.
  • 6SERMANET P,EIGEN D,ZHANG X,et al.Overfeat:integrated recognition,localization and detection using convolutional networks[J].CoRR,2013:abs/1312.6229.
  • 7RUSSAKOVSKY O,DENG J,SU H,et al.Image Net large scale visual recognition challenge[J].CoRR,2014:abs/1409.0575.
  • 8LIN M,CHEN Q,YAN S.Network in network[J].CoRR,2013:abs/1312.4400.
  • 9SUN Y,WANG X,TANG X.Deep learning face representation from predicting 10,000 classes[C]∥Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition.Piscataway,NJ:IEEE,2014:1891-1898.
  • 10TAIGMAN Y,YANG M,RANZATO M A,et al.Deepface:closing the gap to human-level performance in face verification[C]∥Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition.Piscataway,NJ:IEEE,2014:1701-1708.

共引文献2390

同被引文献20

引证文献4

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部