基于图注意力网络的三维网格分割算法  

A 3D mesh segmentation algorithm based on graph attention network

作　　者：李文婷 吴璐璐 周杰 赵勇[1] LI Wenting;WU Lulu;ZHOU Jie;ZHAO Yong(School of Mathematical Sciences,Ocean University of China,Qingdao 266100,Shandong Province,China)

机构地区：[1]中国海洋大学数学科学学院,山东青岛266100

出　　处：《浙江大学学报（理学版）》2023年第6期811-819,共9页Journal of Zhejiang University（Science Edition）

基　　金：山东省自然科学基金资助项目(ZR2018MF006);浙江大学计算机辅助设计与图形系统全国重点实验室开放课题(A2228).

摘　　要：针对三维网格模型分割质量提升问题,提出了感知几何的图注意力网络。首先,定义了感知几何的图注意力系数,利用节点之间的边特征扩展由网络学习得到的注意力系数,引入与节点局部几何信息相关的注意力系数,更好地反映节点之间的相似性。然后,通过调整网络架构,将三维网格模型的几何特征与标签信息共同作为图注意力网络的输入,使标签信息参与网络训练和验证,并通过残差形式的线性连接实现网络的更稳定输出。大量实验结果表明,本文算法能够获得精确的分割边界,其在PSB数据集上的分割准确率较现有经典算法提升约2个百分点,也取得了更好的兰德指数。同时,通过消融实验验证了算法的合理性。Improving the segmentation quality of 3D meshes is always an important problem to computer graphics.To handle this problem,this paper proposes a shape-aware graph attention network.The shape-aware graph attention coefficient is defined to better reflect the similarity between nodes,which not only expands the attention coefficient obtained by network learning with the help of edge features between nodes,but also introduces the attention coefficient related to the local shape information of nodes.On the other hand,the network architecture is adjusted by taking both shape features and labels of 3D mesh model as the input of graph attention network,which enables the participation of labels in network training and verification stages.Residual connection is further employed to make the network output more stable.A large number of experiments show that the proposed algorithm can obtain accurate segmentation boundaries.Compared with the existing classical segmentation algorithms on PSB dataset,the proposed algorithm improves 2%in accuracy,and achieves better Rand index.The reasonableness of the algorithm is proved by ablation experiment.

关 键 词：网格分割 图注意力系数 边特征 局部几何信息 网络架构 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]