http://blog.csdn.net/xgz0124/article/details/50261403
Caffe 作为一款比较流行的DCNN特征提取框架已获得广泛应用。在CVPR/ICCV/ECCV关于DCNN的文章中屡屡出镜。Caffe的安装步骤比较繁琐，但是网上相关的配置文章也有很多，本文就不再啰嗦。
其中基于Python的Caffe特征抽取可参考http://nbviewer.ipython.org/github/BVLC/caffe/blob/master/examples/net_surgery.ipynb
本人由于要对Deep CNN Feature 作进一步的处理，而处理的算法需要在Matlab环境中进行，所以需要Matcaffe提取图像的feature，现在总结了两种方法。
首先是Ross Girshick大神的解决方法，需要修改deploy.prototxt文件，然后直接 scores=net.forward(input_data)，利用forward函数即可得到。该方法要求对deploy.prototxt文件比较熟悉，然后缺点就在于需要修改多次prototxt才能得到不同layer feature。
还有一种方法：caffe_feat=net.blobs('conv1').get_data(),得到4-D single 的矩阵，就为该层(conv1)的特征，caffe_feat(:,:,1,1)为其中一feature map。
另外，MatConvNet也是一个不错的工具，相对于Caffe更适用于Matlab，有兴趣可以尝试下。
——————————————————分割线2016.01.07————————————————————————
Caffe官网 http://caffe.berkeleyvision.org/tutorial/interfaces.html 给出了python、Matlab下的各种接口，
[plain] view plain copy print ? caffe.set_mode_gpu(); caffe.set_device(gpu_id); caffe.set_mode_gpu(); caffe.set_device(gpu_id); [plain] view plain copy print ? net = caffe.Net(model, weights, 'test'); % create net and load weights net = caffe.Net(model, weights, 'test'); % create net and load weights
[plain] view plain copy print ? net.blobs('data').set_data(ones(net.blobs('data').shape)); net.blobs('data').set_data(ones(net.blobs('data').shape));
可以更改各种参数：
[plain] view plain copy print ? net.params('conv1', 1).set_data(net.params('conv1', 1).get_data() * 10); % set weights net.params('conv1', 2).set_data(net.params('conv1', 2).get_data() * 10); % set bias net.layers('conv1').params(1).set_data(net.layers('conv1').params(1).get_data() * 10); net.layers('conv1').params(2).set_data(net.layers('conv1').params(2).get_data() * 10); net.params('conv1', 1).set_data(net.params('conv1', 1).get_data() * 10); % set weights net.params('conv1', 2).set_data(net.params('conv1', 2).get_data() * 10); % set bias net.layers('conv1').params(1).set_data(net.layers('conv1').params(1).get_data() * 10); net.layers('conv1').params(2).set_data(net.layers('conv1').params(2).get_data() * 10);
然后保存网络，可以：
[plain] view plain copy print ? net.save('my_net.caffemodel'); net.save('my_net.caffemodel');
[plain] view plain copy print ? net.blobs('data').set_data(data); net.forward_prefilled(); prob = net.blobs('prob').get_data(); net.blobs('data').set_data(data); net.forward_prefilled(); prob = net.blobs('prob').get_data();
训练：
[plain] view plain copy print ? solver = caffe.Solver('./models/bvlc_reference_caffenet/solver.prototxt'); solver = caffe.Solver('./models/bvlc_reference_caffenet/solver.prototxt'); [plain] view plain copy print ? solver.solve(); solver.solve();
[plain] view plain copy print ? solver.step(1000); iter = solver.iter(); %To get its network: train_net = solver.net; test_net = solver.test_nets(1); solver.restore('your_snapshot.solverstate'); solver.step(1000); iter = solver.iter(); %To get its network: train_net = solver.net; test_net = solver.test_nets(1); solver.restore('your_snapshot.solverstate');
[plain] view plain copy print ? im_data = caffe.io.load_image('./examples/images/cat.jpg'); im_data = imresize(im_data, [width, height]); % resize using Matlab's imresize %or im_data = imread('./examples/images/cat.jpg'); % read image im_data = im_data(:, :, [3, 2, 1]); % convert from RGB to BGR im_data = permute(im_data, [2, 1, 3]); % permute width and height im_data = single(im_data); % convert to single precision im_data = caffe.io.load_image('./examples/images/cat.jpg'); im_data = imresize(im_data, [width, height]); % resize using Matlab's imresize %or im_data = imread('./examples/images/cat.jpg'); % read image im_data = im_data(:, :, [3, 2, 1]); % convert from RGB to BGR im_data = permute(im_data, [2, 1, 3]); % permute width and height im_data = single(im_data); % convert to single precision
Call caffe.reset_all() to clear all solvers and stand-alone nets you have created.
