基于改进级联金字塔网络的人体骨架提取算法  

Human skeleton extraction algorithm based on improved cascaded pyramid network

作　　者：黄友 张娜[1] 包晓安[1] HUANG You;ZHANG Na;BAO Xiao'an(School of Informatics Science and Technology,Zhejiang Sci-Tech University,Hangzhou 310018,China)

机构地区：[1]浙江理工大学信息学院,杭州310018

出　　处：《智能计算机与应用》2021年第7期54-59,共6页Intelligent Computer and Applications

基　　金：国家自然科学基金(620705014,1);浙江省自然科学基金青年基金(LQ20F050010);浙江省重点研发计划项目(2020C03094)。

摘　　要：由于背景复杂和人体容易被遮挡等情况的发生,导致人体骨架关键点的定位精度不高。针对这一问题,本文提出一种基于改进级联金字塔网络的人体骨架提取算法。该算法将注意力模块加入到级联金字塔特征提取网络的每一个残差块之后,根据特征图的不同部分和不同特征图的重要性程度分配不同的权重。同时将原来级联金字塔网络的2次上采样操作改为一次,以减少上采样过程中产生的冗余背景特征。实验结果表明:该算法可以较好地改善原CPN网络在遮挡、背景复杂等情况下定位不精准的问题。Due to the complex background and the easy occlusion of the human body,the positioning accuracy of the key points of the human skeleton is not high.Aiming at this problem,this paper proposes a human skeleton extraction algorithm based on an improved cascaded pyramid network.The algorithm adds the attention module to each residual block of the cascaded pyramid feature extraction network,and assigns different weights according to different parts of the feature map and the importance of different feature maps.At the same time,the two upsampling operations of the original cascaded pyramid network are changed to one to reduce the redundant background features generated in the upsampling process.Experimental results show that the algorithm can better improve the problem of inaccurate positioning in the original CPN network under occlusion and complex background conditions.

关 键 词：级联金字塔网络 注意力机制 多层次特征提取 特征融合 困难关键点挖掘 目标关键点相似度 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]