DeCAF: A deep convolutional activation feature for generic visual recognition

来源:互联网 发布:功夫英语 知乎 编辑:程序博客网 时间:2024/05/01 18:27

DeCAF: An open source implementation of these deep convolutional activation features + all associated networkparameters

               Deep / Layered compositional architectures --->capture salient aspect of a given domain. 

               Deep unsupervised model  outperform (》) gradient histogram features. 


0 0
原创粉丝点击