DL-SFA: Deeply-Learned Slow Feature Analysis for Action Recognition(泛读)
来源:互联网 发布:h3c端口聚合配置s3700 编辑:程序博客网 时间:2024/06/05 08:04
一.文献名字和作者
DL-SFA: Deeply-Learned Slow Feature Analysis for Action Recognition,Lin Sun, Kui Jiay, Tsung-Han Chany, Yuqiang Fang, Gang Wang], Shuicheng Yan
二.阅读时间
2014年9月30日
三.文献的贡献点
这篇文章主要是将用于视频中行为识别的SFA与深度学习进行结合,虽然说是结合,其实并没有用到深度学习的训练过程,只是将两个SFA进行级联,不过是在级联的过程中,将第一层SFA的核作为3D卷积层的卷积核对视频进行卷积,然后使用了两个maxpooling层,然后再次进行了SFA的学习。不知道为什么这样做可行,可能是对因为于这个领域没有多大研究的。
个人感觉这篇文章和深度学习其实关系并不是特别大,主要还是在于将SFA进行了级联,特点在于级联的方式。
版权所有,欢迎转载,转载请注明出处,谢谢
0 0
- DL-SFA: Deeply-Learned Slow Feature Analysis for Action Recognition(泛读)
- Slow Feature Analysis(SFA)模型
- 《Deeply-Learned Feature for Age Estimation》论文阅读笔记
- A Theoretical Analysis of Feature Pooling in Visual Recognition(泛读)
- Deeply Learned Attributes for Crowded Scene Understanding
- 论文选读Range-Sample Depth Feature for Action Recognition
- 论文选读Range-Sample Depth Feature for Action Recognition
- DeepID2+:Deeply Learned Attributes for Crowded Scene Understanding
- Deeply Learned Attributes for Crowded Scene Understanding(WWW dataset)
- Selective Search for Object Recognition(泛读)
- Journals for Action Recognition
- Combining the Right Features for Complex Event Recognition(泛读)
- 人群场景的属性--Deeply Learned Attributes for Crowded Scene Understandin
- 论文笔记Deeply-Learned Part-Aligned Representations for Person Re-Identification
- A Theoretical Analysis of Feature Pooling in Visual Recognition
- Deep Feature Flow for Video Recognition读书笔记
- Deep Feature Flow for Video Recognition 安装
- Discriminative Learned Dictionaries for Local Image Analysis学习笔记
- ABAP FIELD-SYMBOLS 有大作用- 将没有可改参数的增强出口变得也能改主程序的值了
- WordPress时间函数the_tiem()详解
- oracle如何干预执行计划-使用hints
- 优秀的开源库
- wav格式 常识
- DL-SFA: Deeply-Learned Slow Feature Analysis for Action Recognition(泛读)
- 基于二叉树SVM多类分类算法研究
- 推荐一个非常不错的Linux资源网站
- How will the new iPhone screen sizes affect iOS developers?
- MFC几个关于字符串处理的全局函数
- Oracle判断值为非数字
- cannot simultaneously fetch multiple bags
- SVN服务器搭建和使用(三)
- 探秘腾讯Android手机游戏平台之不安装游戏APK直接启动法