首页|DFE3D:双重特征增强的三维点云类增量学习

DFE3D:双重特征增强的三维点云类增量学习

扫码查看
原文链接
  • NETL
  • NSTL
  • 万方数据
随着点云采集技术的发展和三维应用需求的增加,实际场景要求针对流动数据持续动态地更新点云分析网络.对此,提出了双重特征增强的三维点云类增量学习方法,通过增量学习使点云目标分类技术能够适应新数据中不断出现新类别目标的场景.该方法通过对点云数据特性和旧类信息的研究分别提出了差异性局部增强模块和知识注入网络,以缓解类增量学习中的新类偏好问题.具体而言,差异性局部增强模块通过感知丰富的局部语义,表征出三维点云物体中不同的局部结构特性.随后,根据目标中每个局部结构的全局信息获得各个局部的重要性权重,强化对差异性局部特征的感知,从而提高新旧类特征差异性.另外,知识注入网络将旧模型中的旧知识注入新模型的特征学习过程中,增强后的混合特征能够更有效缓解旧类信息不足导致的新类偏好加剧现象.在三维点云数据集ModelNet40,ScanObjectNN,ScanNet,ShapeNet上的实验表明,该方法与现有最优方法相比,在 4 个数据集上的平均增量准确率有2.03%、2.18%、1.65%、1.28%提升.
DFE3D:Class-incremental Learning for 3D Point Cloud with Dual Feature Enhancement
As point cloud acquisition technology develops and the demand for 3D applications increases,real-world scenarios require continuous and dynamic updating of the point cloud analysis network with streaming data.This study proposes a dual feature enhancement for the class-incremental 3D point cloud object learning method,which adapts point cloud object classification to scenarios where new category objects keep emerging in newly acquired data through incremental learning.This study proposes a discriminative local enhancement module and knowledge injection network respectively to alleviate new class bias problems in class-incremental learning by studying the characteristics of point cloud data and old class information.Specifically,the discriminative local enhancement module characterizes the various local structural characteristics of 3D point cloud objects by perceiving expressive local features.Subsequently,the importance weights of each local structure are obtained based on the global information of each local structure,enhancing the perception of differential local features and improving the differentiation of new and old class features.Furthermore,the knowledge injection network injects old knowledge from the old model into the feature learning process of the new model.The enhanced hybrid features can effectively mitigate the increased new class bias caused by the lack of old class information.Under the incremental learning experimental settings of the 3D point cloud datasets ModelNet40,ScanObjectNN,ScanNet,and ShapeNet,extensive experiments show that compared with existing state-of-art methods,the method in this study has an average incremental accuracy improvement of 2.03%,2.18%,1.65%,and 1.28%on the four datasets.

3D point cloud object classificationclass-incremental learningdiscriminative local enhancementknowledge injectioncatastrophic forgettingpoint cloud representation

孙昊、帅惠、许翔、刘青山

展开 >

南京信息工程大学计算机学院,南京 210044

南京邮电大学计算机学院、软件学院、网络空间安全学院,南京 210023

南京航空航天大学,南京 211106

南京邮电大学,南京 210023

展开 >

三维点云目标分类 类增量学习 差异性局部增强 知识注入 灾难性遗忘 点云表征

科技创新2030—"新一代人工智能"重大项目

2021ZD0112200

2024

10.15888/j.cnki.csa.009574

计算机系统应用
中国科学院软件研究所

计算机系统应用

CSTPCD
影响因子：0.449
ISSN：1003-3254
年,卷(期)：2024.33(8)