[1]钟 诚,周浩杰,韦海亮.一种基于注意力机制的三维点云物体识别方法[J].计算机技术与发展,2020,30(04):41-45.[doi:10. 3969 / j. issn. 1673-629X. 2020. 04. 008]
　ZHONG Cheng,ZHOU Hao-jie,WEI Hai-liang.A 3D Point Cloud Object Recognition Method Based on Attention Mechanism[J].COMPUTER TECHNOLOGY AND DEVELOPMENT,2020,30(04):41-45.[doi:10. 3969 / j. issn. 1673-629X. 2020. 04. 008]
点击复制

一种基于注意力机制的三维点云物体识别方法()
分享到：

《计算机技术与发展》[ISSN:1006-6977/CN:61-1281/TN]

卷:
30
期数:
2020年04期
页码:
41-45
栏目:
智能、算法、系统工程
出版日期:
2020-04-10

文章信息/Info

Title:
A 3D Point Cloud Object Recognition Method Based on Attention Mechanism
文章编号:
1673-629X(2020)04-0041-05
作者:
钟 诚; 周浩杰; 韦海亮
数学工程与先进计算国家重点实验室,江苏 无锡 214000
Author(s):
ZHONG Cheng; ZHOU Hao-jie; WEI Hai-liang
State Key Laboratory of Mathematical Engineering and Advanced Computing,Wuxi 214000,China
关键词:
注意力机制; 点云; 物体识别; 池化; 稀疏卷积
Keywords:
attention mechanism; point cloud; object recognition; pooling; sparse convolution
分类号:
TP301
DOI:
10. 3969 / j. issn. 1673-629X. 2020. 04. 008
摘要:
三维点云数据通常具备无序排列的结构。 在三维点云数据处理领域,深度学习模型通常会利用最大池化等对称操作来处理点云的排列不变性。最大池化方法一方面会破坏点云的信息结构,使得局部信息与全局信息难以交互。另一方面,最大池化方法对点云信息过度压缩,得到的特征对局部细节描述不足。 针对上述问题,提出了 AttentionPointNet 的网络结构。 该网络利用注意力机制,使每个点与点云其余部分进行特征交互,实现了局部与全局信息的综合。为降低最大池化造成的信息损失,提出了一种稀疏卷积方法来替代池化操作。这种方法利用大步长的稀疏卷积实现全局信息的提取。在 ModelNet40 数据集上,AttentionPointNet 取得了87.2% 的准确率。不使用池化层,完全采用卷积层实现的模型取得了86.2% 的分类准确率。
Abstract:
3D point cloud data usually has an unordered structure. In the field of point cloud data processing,deep learning models usually use the symmetry operations such as maximum pooling to deal with the permutation invariance of point clouds. On the one hand, this approach often destroys local information of point cloud data. On the other hand,the maxpooling method over-compresses point cloud in-formation,and the extracted features are insufficiently described for local details. Aiming at those problems, we propose a network structure called AttentionPointNet which uses the attention mechanism to make each point interact with the rest of the point cloud to achieve the integration of local and global information. In order to reduce the information loss caused by the maximum pooling, we propose a sparse convolution to replace the pooling layer,which uses large stride sparse convolution to extract global information. On the ModelNet40 dataset,AttentionPointNet achieves 87.2% classification accuracy. The model,which only uses convolution layers to replace maxpooling layer,achieves 86.2% classification accuracy.

相似文献/References:

[1]赵越超 李忠科 王勇 吕培军.基于OpenGL的三维牙颌模型可视化研究[J].计算机技术与发展,2008,(01):119.
　ZHAO Yue-chao,LI Zhong-ke,WANG Yong,et al.Research of 3D Dental Visualization Based on OpenGL[J].COMPUTER TECHNOLOGY AND DEVELOPMENT,2008,(04):119.
[2]张琴 蔡勇.支持向量学习机在点云去噪中的应用[J].计算机技术与发展,2011,(06):85.
　ZHANG Qin,CAI Yong.Application of Support Vector Machine in Point Clouds Denoising[J].COMPUTER TECHNOLOGY AND DEVELOPMENT,2011,(04):85.
[3]周春艳 李勇 邹峥嵘.三维点云ICP算法改进研究[J].计算机技术与发展,2011,(08):75.
　ZHOU Chun-yan,LI Yong,ZOU Zheng-rong.Three-Dimensional Cloud ICP Algorithm Improvement[J].COMPUTER TECHNOLOGY AND DEVELOPMENT,2011,(04):75.
[4]谢晓燕,吴锦桥. 一种全自动三维点云配准及比例约束方法[J].计算机技术与发展,2015,25(03):63.
　XIE Xiao-yan,WU Jin-qiao. An Automatic Method of 3 D Point Cloud Registration and Dimension Adjustment[J].COMPUTER TECHNOLOGY AND DEVELOPMENT,2015,25(04):63.
[5]李梦洁,董峦.基于 PyTorch 的机器翻译算法的实现[J].计算机技术与发展,2018,28(10):160.[doi:10.3969/ j. issn.1673-629X.2018.10.033]
　LI Meng-jie,DONG Luan.Implementation of Machine Translation Algorithm Based on PyTorch[J].COMPUTER TECHNOLOGY AND DEVELOPMENT,2018,28(04):160.[doi:10.3969/ j. issn.1673-629X.2018.10.033]
[6]李东欣,禹 龙,田生伟,等.注意力机制的 LSTM-DBN 维语人称代词指代消解[J].计算机技术与发展,2019,29(07):33.[doi:10. 3969 / j. issn. 1673-629X. 2019. 07. 007]
　LI Dong-xin,YU Long,TIAN Sheng-wei,et al.Attention Mechanism of LSTM-DBN Uyghur Personal Pronoun Anaphora Resolution[J].COMPUTER TECHNOLOGY AND DEVELOPMENT,2019,29(04):33.[doi:10. 3969 / j. issn. 1673-629X. 2019. 07. 007]
[7]尹 鹏,周 林,郭 强,等.基于短语级注意力机制的关系抽取方法[J].计算机技术与发展,2019,29(09):24.[doi:10. 3969 / j. issn. 1673-629X. 2019. 09. 005]
　YIN Peng,ZHOU Lin,GUO Qiang,et al.Relation Extraction Based on Phrase-level Attention[J].COMPUTER TECHNOLOGY AND DEVELOPMENT,2019,29(04):24.[doi:10. 3969 / j. issn. 1673-629X. 2019. 09. 005]
[8]廖中平,白慧鹏,陈 立.基于双边滤波改进的点云平滑算法[J].计算机技术与发展,2019,29(11):42.[doi:10. 3969 / j. issn. 1673-629X. 2019. 11. 009]
　LIAO Zhong-ping,BAI Hui-peng,CHEN Li.Improved Denoising of Point-sampled Model Based on Bilateral Filtering[J].COMPUTER TECHNOLOGY AND DEVELOPMENT,2019,29(04):42.[doi:10. 3969 / j. issn. 1673-629X. 2019. 11. 009]
[9]刘保安.ICP 算法在文物三维重建中的应用[J].计算机技术与发展,2020,30(01):217.[doi:10. 3969 / j. issn. 1673-629X. 2020. 01. 039]
　LIU Bao-an.Application of ICP Algorithm in Three-dimensional Reconstruction of Cultural Relics[J].COMPUTER TECHNOLOGY AND DEVELOPMENT,2020,30(04):217.[doi:10. 3969 / j. issn. 1673-629X. 2020. 01. 039]
[10]王乾铭,李 吟.基于深度学习的个性化聊天机器人研究[J].计算机技术与发展,2020,30(04):79.[doi:10. 3969 / j. issn. 1673-629X. 2020. 04. 015]
　WANG Qian-ming,LI Yin.Research on Personalized Chatbot Based on Deep Learning[J].COMPUTER TECHNOLOGY AND DEVELOPMENT,2020,30(04):79.[doi:10. 3969 / j. issn. 1673-629X. 2020. 04. 015]

更新日期/Last Update: 2020-04-10