| 动作识别算法的评估策略探讨 |
| |
| 引用本文： | 高赞,张桦,蔡安妮.动作识别算法的评估策略探讨[J].光电子．激光,2012(6):1166-1172. |
| |
| 作者姓名： | 高赞 张桦 蔡安妮 |
| |
| 作者单位： | 计算机视觉与系统省部共建教育部重点实验室,天津理工大学;天津市智能计算及软件新技术重点实验室,天津理工大学;北京邮电大学信息与通信工程学院;计算机视觉与系统省部共建教育部重点实验室,天津理工大学;天津市智能计算及软件新技术重点实验室,天津理工大学;北京邮电大学信息与通信工程学院 |
| |
| 基金项目： | 国家自然科学基金(90920001);天津市科技支撑计划(10ZCKFGX00400)资助项目 |
| |
| 摘 要： | 以时空兴趣点特征和支持向量机(SVM)分类器识别方法为基本算法,在广泛使用的公开动作数据集KTH上,从不同角度考察评估策略对动作识别算法性能的影响。实验表明,当采用不同的交叉实验方法时,算法性能的波动最大达到10.5%,而不同数据集划分方法对算法性能的影响则达到11.87%。因此,通过量化分析得出的结论,可以充分地比较现有算法的真实差异,并为设计合理的评估策略提供参考。
|
| 关 键 词： | 时空特征 支持向量机(SVM) 动作识别 评估策略 |
Discussion on the assessment strategy of action recognition algorithms |
| |
| GAO Zan,ZHANG Hua and CAI An-ni.Discussion on the assessment strategy of action recognition algorithms[J].Journal of Optoelectronics·laser,2012(6):1166-1172. |
| |
| Authors: | GAO Zan ZHANG Hua and CAI An-ni |
| |
| Affiliation: | Key Laboratory of Computer Vision and System,of Tianjin and Ministry of Education Tianjin University of Technology Tian-jin 300384,China;Tianjin Key Laboratory of Intelligence Computing and Novel Software Technology,Tianjin Univers-ity of Technology,Tianjin 300384,China;School of Information and Communication Engineering,Beijing University of Posts and Telecommunications,Beijing 100876,Cahina;Key Laboratory of Computer Vision and System,of Tianjin and Ministry of Education Tianjin University of Technology Tian-jin 300384,China;Tianjin Key Laboratory of Intelligence Computing and Novel Software Technology,Tianjin Univers-ity of Technology,Tianjin 300384,China;School of Information and Communication Engineering,Beijing University of Posts and Telecommunications,Beijing 100876,Cahina |
| |
| Abstract: | Action recognition is a hot research topic,but the performance assessment strategies of algorithms have not had an accepted practice.In this paper,we adopt spatio-temporal features and support vector machine(SVM) model as our action recognition algorithm,and then well assess the effect of different assessment strategies on our action recognition algorithm in widely used public dataset KTH.Experimental results show that when different cross-experimental methods are employed,the performance fluctuation of algorithms reaches 10.5%.And when different division methods for KTH datasets are used,the performance fluctuation of algorithms gets 11.87%.Thus,according to conclusions in this paper,we can find the real difference among existing algorithms,and supply the reference for designing reasonable assessment strategy. |
| |
| Keywords: | spatio-temporal features support vector machine(SVM) action recognition assessment strategy |
| 本文献已被 CNKI 等数据库收录！ |
| 点击此处可从《光电子．激光》浏览原始摘要信息 |
|
点击此处可从《光电子．激光》下载全文 |
|
