基于知识共享的遮挡人体姿态估计网络  

Occluded human pose estimation network based on knowledge sharing

作　　者：江佳鸿 夏楠 李长吾[1] 于鑫淼 JIANG Jiahong;XIA Nan;LI Changwu;YU Xinmiao(School of Information Science and Engineering,Dalian Polytechnic University,Dalian 116034,China)

机构地区：[1]大连工业大学信息科学与工程学院,辽宁大连116034

出　　处：《浙江大学学报（工学版）》2024年第10期2001-2010,共10页Journal of Zhejiang University：Engineering Science

基　　金：教育部产学合作协同育人资助项目(220603231024713).

摘　　要：现有人体姿态估计方法处理遮挡情况时性能较差,为此提出新的估计网络,包含遮挡区域强化卷积网络(OCNN)和遮挡特征补偿图卷积网络(OGCN).设计高低阶特征匹配注意力以强化遮挡区域特征,由OCNN提取高适配权重,通过少量遮挡数据的方式实现遮挡部位的强化检测.由OGCN消除障碍物特征,通过强化关键点共有及专有属性的方式补偿节点特征;进行邻接矩阵重要性加权以改善遮挡部位特征质量,提升检测精度.所提网络在数据集COCO2017、COCO-Wholebody、CrowdPose上的检测精度分别为78.5%、67.1%、77.8%,优于对比算法.在自建遮挡数据集上所提网络节约了75%的训练数据使用.A new estimation network was proposed for improving the insufficient occlusion handling ability of existing human pose estimation methods.An occluded parts enhanced convolutional network(OCNN)and an occluded features compensation graph convolutional network(OGCN)were included in the proposed network.A high-low order feature matching attention was designed to strengthen the occlusion area features,and highadaptation weights were extracted by OCNN,achieving enhanced detection of the occluded parts with a small amount of occlusion data.OGCN strengthened the shared and private attribute compensation node features by eliminating the obstacle features.The adjacency matrix was importance-weighted to enhance the quality of the occlusion area features and to improve the detection accuracy.The proposed network achieved detection accuracy of 78.5%,67.1%,and 77.8%in the datasets COCO2017,COCO-Wholebody,and CrowdPose,respectively,outperforming the comparative algorithms.The proposed network saved 75%of the training data usage in the selfbuilt occlusion dataset.

关 键 词：人体姿态估计 遮挡处理 高低阶特征匹配 节点特征补偿 邻接矩阵加权 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]