摘要
近年来,人体活动识别技术得到了迅速发展,并在安全防御、医疗健康、智能交互、体育运动等领域得到广泛应用。本文结合国内外的相关文献,围绕人体活动识别,对知识图谱、人体活动数据集及人体活动识别方法及应用进行综述。首先,指出了不同数据采集方式的人体活动识别数据集的特点。其次,梳理了各种不断提高模型的准确度和性能的人体活动识别具体方法。再次,指出了知识图谱在人体识别活动中发挥的揭示实体间的关系,为人体活动解析提供隐含知识等重要作用。最后,面对人体活动识别的实际应用,对当前研究存在的问题和发展趋势进行了总结和展望。本研究为人体活动识别研究的技术发展和实际应用提供了方向指导和技术建议。
In recent years,the technology for recognizing human activity has progressed rapidly and has found widespread applications in the fields of security,healthcare,human-computer interaction,and sports.Based on the relevant literature at home and abroad,this paper reviews the knowledge graph,human activity data sets and human activity recognition methods and applications.It begins by identifying the distinguishing features of human activity recognition datasets collected using various methods.Subsequently,it examines specific techniques aimed at enhancing the accuracy and performance of human activity recognition models.Furthermore,it underscores the crucial role of knowledge graphs in unveiling the relationships between entities in human activity recognition and in providing implicit knowledge for the analysis of human activity.Finally,in light of the practical implementation of human activity recognition,it summarizes existing research challenges and anticipates future development trends.This study provides direction and technical suggestions for the technology development and practical application of human activity recognition research.
作者
何玲
杨一
刘丹
章杰
HE Ling;YANG Yi;LIU Dan;ZHANG Jie(Key Laboratory of Advanced Manufacturing Technology of the Ministry of Education,Guizhou University,Guiyang 550025,China;College of Computer Science and Technology,Guizhou University,Guizhou Guiyang 550025,China)
出处
《贵州大学学报（自然科学版）》
2024年第5期49-57,共9页
Journal of Guizhou University:Natural Sciences
基金
贵州省科技支撑计划项目(黔科合支撑【2021】一般439,黔科合支撑【2023】一般124)。
关键词
知识图谱
人体活动数据集
人体活动识别
knowledge graph
human activity data sets
human activity recognition
