| 具有多峰正态分布属性的视频语义分类研究 |
| |
| 引用本文： | 魏维,徐建,刘凤玉,许满武.具有多峰正态分布属性的视频语义分类研究[J].计算机科学,2006,33(4):111-114. |
| |
| 作者姓名： | 魏维 徐建 刘凤玉 许满武 |
| |
| 作者单位： | 1. 南京理工大学计算机科学与技术系,南京,210094
2. 南京大学计算机科学与技术系,南京,210008 |
| |
| 基金项目： | 中国科学院资助项目;江苏省教育厅科研项目 |
| |
| 摘 要： | 视频语义分类中常遇到多峰正态分布属性，如采用单峰值正态分布设计的贝叶斯分类模型会造成较大分类误差。本文采用定步长组合划分算（FLCPA）对多峰分布属性值域按类进行划分，以留一校验法（LOOCV）估算分类错误，找出给定步长下属性的多峰分布边界点，并用监督参数估计推断出每个分段区间上的概率分布函数，从而得到整个值域上的总体分布。此外，文中给出了涉及多峰分布属性的视频语义分类器设计步骤。实验数据表明，该方法能明显降低分类错误，有效提高分类性能。
|
| 关 键 词： | 贝叶斯分类器 多峰正态分布属性 视频语义分类 留一校验 类条件概率密度函数 |
Study of Video Semantic-concept Classifier with Multi-normal Distribution Attribute |
| |
| WEI Wei,XUE Jian,LIU Feng-Yu,XU Man-Wu.Study of Video Semantic-concept Classifier with Multi-normal Distribution Attribute[J].Computer Science,2006,33(4):111-114. |
| |
| Authors: | WEI Wei XUE Jian LIU Feng-Yu XU Man-Wu |
| |
| Affiliation: | 1.Department of Computer Science and Technology, Nanjing University of Science and Technology, Nanjing 210094;2. Department of Computer Science and Technology, Nanjing University, Nanjing 210008 |
| |
| Abstract: | Attributes with multi-normal distribution are common in classifier design for video-semantic concept.In this case,a model assuming that the value of attributes for each class is normally distributed with some mean will lead to poor classification performance.In the paper,an approach based on fixed-length combination partition algorithm(FLC- PA)is presented in the partition of attribute value-field.Leave-one-out cross-validation(LOOCV)is used to estimate classifier error.In addition,the detail of classifier design about multi-normality distribution attribute is given.The re- suit of experiment indicate the method could reduce classifier error and improve classifier performance. |
| |
| Keywords: | Bayesian classifier Multi normal distribution Semantic-based video classifier LOOCV Class-conditionaldensity function |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
|
点击此处可从《计算机科学》下载全文 |
