为什么要学习Caffe呢?因为通过Caffe可以综合性的学习Deep Learning,CUDA,Spark(CaffeOnSpark),cuDNN,OpenBLAS,MKL,C++,Python,Linux等知识。安装Caffe之前主要是安装CUDA,BLAS等,关于cuDNN和Python是可以选择的。Caffe详细的安装过程参考[1][2]。
Caffe(Convolutional Architecture for Fast Feature Embedding)是什么呢?简单来说Caffe主要是一个关于CNN的深度学习框架,可以用于模式识别,图形图像,机器视觉等应用。
1. Caffe的目录结构
参考文献:
[1] Setting up a Deep Learning Machine from Scratch (Software):https://github.com/saiprashanths/dl-setup#caffe
[2] Caffe:http://caffe.berkeleyvision.org/
[3] Caffe:https://github.com/BVLC/caffe
[4] 知乎Caffe:http://www.zhihu.com/search?type=content&q=Caffe
[5] Caffe —— Deep learning in Practice:http://blog.csdn.net/abcjennifer/article/details/46424949
[6] Caffe深度学习框架上手教程:http://www.open-open.com/lib/view/open1421995285109.html
[7] Caffe Demos:http://demo.caffe.berkeleyvision.org/
[8] Yangqing Jia (贾扬清):http://daggerfs.com/