基于深度学习的2D人体姿态估计浅析  

Analysis of 2DHuman Pose Estimation Based on Deep Learning

作　　者：任肖月 王俊英[1,2,3] 陈晗晗 REN Xiaoyue;WANG Junying;CHEN Hanhan(Hubei Key Laboratory of Intelligent Vision Based Monitoring for Hydroelectric Engineering,China Three Gorges University,Yichang 443002,China;Hubei Province Engineering Research Center for Building Quality Inspection Equipment,Three Gorges University,Yichang 443002,China;College of Computer and Information Technology,China Three Gorges University,Hubei,Yichang 443002,China)

机构地区：[1]三峡大学水电工程智能视觉监测湖北省重点实验室 [2]三峡大学湖北省建筑质量检测装备工程技术研究中心 [3]三峡大学计算机与信息学院,湖北宜昌443002

出　　处：《长江信息通信》2024年第1期40-43,47,共5页Changjiang Information & Communications

摘　　要：近年来,随着深度学习技术的发展,基于深度学习的2D人体姿态估计展现出了极其优异的效果。文章主要对2014年以来基于深度学习的2D人体姿态估计相关内容进行了分析和评述,包括对主流方法的介绍和对数据集评价指标的总结对比,最后还讨论了2D人体姿态估计的未来研究方向。In recent years,with the development of deep learning technology,2D human pose estimation based on deep learning has shown excellent results.This paper mainly describes the relevant content of 2D human pose estimation based on deep learning since 2014,including the introduction of mainstream methods and the summary and comparison of dataset evaluation indicators,and finally looks forward to the future of human pose estimation based on our own understanding.

关 键 词：人体姿态估计 深度学习 关键点检测 神经网络 

分 类 号：TP301[自动化与计算机技术—计算机系统结构]