中国计量大学学报2017，Vol.28Issue(1)：57-62,6.DOI:10.3969/j.issn.2096-2835.2017.01.010
基于结构化学习的小群体主导者检测方法
An leadership detection algorithm in small groups based on structural learning
摘要
Abstract
Crowd analysis is an interesting theme in pattern recognition and machine learning.In particular,the opportunity of leadership detection in small groups opens up to new perspectives in video surveillance and other computer vision fields.Meanwhile,leadership detection in small groups is also an important part of crowd analysis.In this paper,we proposed a structural SVM learning framework and a new feature model based on pedestrians,position analysis and time-lag analysis.Results show that the method is effective in the dataset of crowd analysis.关键词
人群分析/主导者检测/时间滞后分析/结构化SVMKey words
crowd analysis/leader identification/time-lag analysis/structural SVM分类
信息技术与安全科学引用本文复制引用
章东平,钱乐义..基于结构化学习的小群体主导者检测方法[J].中国计量大学学报,2017,28(1):57-62,6.基金项目
浙江省自然科学基金资助项目(No.LY15F020021),浙江省公益性项目(No.2016C31079). （No.LY15F020021）
