二维码（扫一下试试看！） | | 结合张量特征和孪生支持向量机的群体行为识别 | | Group Activity Recognition Based on Tensor Features and Twin Support Vector Machines | | 投稿时间：2018-07-29 | | DOI：10.15918/j.tbit1001-0645.2019.10.012 | | 中文关键词:群体行为识别张量特征孪生支持向量机粒子群优化 | | English Keywords:group activity recognitiontensor featuretwin support vector machineparticle swarm optimization | | 基金项目:国家自然科学基金资助项目（61672032）；安徽省重点实验室开放课题资助项目（2016-KFKT-003） | | | 摘要点击次数:574 | | 全文下载次数:309 | | 中文摘要: | | 给出一种结合张量特征和孪生支持向量机的群体行为识别算法，以提高对视频中群体行为识别的准确率.首先通过群成员关节点骨架的姿态结构信息和群成员的社会网络信息描述群体在每一帧中的行为，并采用张量形式表示；然后使用多路非线性特征映射分解张量核，并利用粒子群优化张量核孪生支持向量机的模型参数；最后结合张量特征和孪生支持向量机实现视频中的群体行为识别.CAD2数据集和自建数据集上的实验结果表明，张量特征能够有效地表示群体行为，相比经典算法，所提算法能有效提高群体行为识别的准确率. | | English Summary: | | To improve the accuracy of group activity recognition in video, a group activity recognition algorithm was proposed based on tensor feature and twin support vector machine. Firstly, the activity of group in each frame was described by combining the posture structure information in the joint skeleton of the group members and the social network information of the group. The tensor form was used to represent the features of group activity. Then, the tensor kernel was decomposed by using multi-channel nonlinear feature mapping and the model parameters of the tensor kernel twin support vector machine were optimized by using the particle swarm optimization method. Finally, the group activity recognition in video was realized by combining tensor features and twin support vector machine. Experiments performed on the CAD2 dataset and the self-built dataset show that the tensor feature can effectively represent the group activity. Compared with the existing approach, the proposed algorithm can effectively improve the accuracy of the group activity recognition. | | 查看全文查看/发表评论下载PDF阅读器 | |
余斌,李晓风,赵赫.基于区块链存储扩展的结构化数据管理方法[J].北京理工大学学报（自然科学版）,2019,39(11):1160~1166.YUBin,LIXiao-feng,ZHAOHe.StructuredDataManagementMethodBasedonScalableBlockchai ... 北京理工大学科研学术 本站小编 Free考研考试 2021-12-21罗玉文,柳丹,范雄华,江晶.空地概率数据关联转换量测滤波算法[J].北京理工大学学报（自然科学版）,2019,39(12):1268~1276.LUOYu-wen,LIUDan,FANXiong-hua,JIANGJing.StudyonAir-GroundProbabilityDataAssoci ... 北京理工大学科研学术 本站小编 Free考研考试 2021-12-21李金,刘文杰,孟祥莲,边晨源,梁洪.脑影像与基因组数据关联的交互可视化方法研究[J].北京理工大学学报（自然科学版）,2019,39(S1):12~18.LIJin,LIUWen-jie,MENGXiang-lian,BIANChen-yuan,LIANGHong.ResearchonInterac ... 北京理工大学科研学术 本站小编 Free考研考试 2021-12-21计卫星,张露露,陈娟,邹天刚,罗辉,郭静,高志伟.一种时序数据多项式拟合加速方法[J].北京理工大学学报（自然科学版）,2018,38(5):519~524.JIWei-xing,ZHANGLu-lu,CHENJuan,ZOUTian-gang,LUOHui,GUOJing,GAOZhi-wei.P ... 北京理工大学科研学术 本站小编 Free考研考试 2021-12-21贺秦禄,边根庆,邵必林,叶娜.一种基于云存储系统的自适应数据碎片恢复优化方法[J].北京理工大学学报（自然科学版）,2018,38(8):841~847.HEQin-lu,BIANGen-qing,SHAOBi-lin,YENa.AnAdaptiveDefragmentationRestoreOpt ... 北京理工大学科研学术 本站小编 Free考研考试 2021-12-21闫书法,马彪,郑长松,王立勇,朱礼安,马源.基于劣化数据的综合传动装置剩余寿命预测[J].北京理工大学学报（自然科学版）,2018,38(11):1126~1133.YANShu-fa,MABiao,ZHENGChang-song,WANGLi-yong,ZHULi-an,MAYuan.Remain ... 北京理工大学科研学术 本站小编 Free考研考试 2021-12-21徐小龙,龚培培.面向云端融合计算的防御性数据销毁机制[J].北京理工大学学报（自然科学版）,2017,37(3):281~286.XUXiao-long,GONGPei-pei.DefensiveDataDestructionMechanismforCloud-P2PComputing[J].Tra ... 北京理工大学科研学术 本站小编 Free考研考试 2021-12-21郑凤,陈艺戬.基于多径的双极化信道信息反馈方法[J].北京理工大学学报（自然科学版）,2017,37(4):365~370.ZHENGFeng,CHENYi-jian.CSIFeedbackBasedonMulti-PathsInformationinDual-PolarizedMIMOSystem ... 北京理工大学科研学术 本站小编 Free考研考试 2021-12-21李向荣,李帅,王国辉,李国栋,李帆.基于有限元数据交会的部件毁伤评估方法研究[J].北京理工大学学报（自然科学版）,2017,37(4):430~435.LIXiang-rong,LIShuai,WANGGuo-hui,LIGuo-dong,LIFan.DamageAssessmentMethodo ... 北京理工大学科研学术 本站小编 Free考研考试 2021-12-21张真宁,孙华飞,韩希武,曹丽梅.晶粒直径的信息几何结构[J].北京理工大学学报（自然科学版）,2017,37(4):436~440.ZHANGZhen-ning,SUNHua-fei,HANXi-wu,CAOLi-mei.TheInformationGeometricStructuresoftheS ... 北京理工大学科研学术 本站小编 Free考研考试 2021-12-21
| |
