Multiple People Tracking by Lifted Multicut and Person Re-identification
来源:互联网 发布:淘宝阿迪达斯旗舰店 编辑:程序博客网 时间:2024/05/16 12:20
来源:CVPR2017
创新点:
通过提升多维数据和行人重识别来进行多人跟踪。在本文中,提出了一种新颖的基于图的公式,用以随着时间的推移通过解决一个最小化成本(cost)提升维度(lifted multicut)的问题的示例来连接并聚集(cluster)行人假设。本文的模型通过介绍一种机制概括了之前的工作。这使得可以在一定程度上奖励那些把具有相似外观的检测分配给同一行人的跟踪结果。为有效地匹配具有更长时间间隔的假设,开发出一种新的深层架构用以行人重识别。该架构联合了用深度网络提取的整体特征和用最先进的姿态估计模型所获得的身体姿态布局。
贡献:
①通过融合行人的姿态信息,设计并训练了行人重识别的深度网络,这提供了一种连接行人假设的机制。
②提出把多人跟踪看作是the minimum cost lifted multicut problem。本文为跟踪图介绍两种类型的边(regular and lifted edges)。Regular edges定义了图中的可行解集合,也就是说确定哪一对节点应该被joint/cut。The lifted edges添加了额外的长距离信息,在不修改可行解集合的情况下,判断哪个节点应该被joint/cut。本文方法可以编码长距离信息,但是会以一种统一且严格的方式在可行解中通过强制有效路径来惩罚长期的错误联合(如,外观相似的人)。
阅读全文
1 0
- Multiple People Tracking by Lifted Multicut and Person Re-identification
- Multiple People Tracking by Lifted Multicut and Person Re-identification
- CVPR 2017 Multiple People Tracking by Lifted Multicut and Person Re-identification 阅读笔记
- Re-ID: Person Re-identification by Local Maximal Occurrence Representation and Metric Learning 论文解析
- 【Person Re-ID】Person Re-identification: Past, Present and Future
- Person Re-Identification by Iterative Re-Weighted Sparse Ranking
- Person re-identification by Local Maximal Occurrence representation and metric learning
- paper reading——《Improving Person Re-identification by Attribute and Identity Learning》
- Person Re-identification Datasets
- Person Re-identification Datasets
- Person Re-identification Overview
- Person Re-identification Overview
- Person Re-identification Datasets
- 一,survey:person re-identification:past,present and future
- 论文笔记-Person Re-identification Past, Present and Future
- 论文笔记-Person Re-identification Past, Present and Future
- 论文笔记-Person Re-identification Past, Present and Future
- ICCV2017: Unlabeled Samples Generated by GAN Improve the Person Re-Identification Baseline in Vitro
- Spark中算子
- TPS波动各种原因总结,做过的项目大概总结下有5中以上情况...
- 树莓派(RPi) CentOS7扩展内存卡剩余空间
- 【ML学习笔记】10:机器学习中的数学基础7(张集,秩,维度,行阶梯形)
- 不过的小东东
- Multiple People Tracking by Lifted Multicut and Person Re-identification
- 汇编冒泡排序
- spring 核心与源码解析(1):IoC之BeanFactory
- EOJ Monthly 2017.12
- leetcode题解-86. Partition List
- 欢迎使用CSDN-markdown编辑器
- 跟随鼠标移动
- 二叉树 深度优先搜索(DFS)、广度优先搜索(BFS)
- 线性约束最小方差准则