|
环境:windows 7+matlab2016a+vs2013
caffe下载地址:https://github.com/BVLC/caffe/tree/windows
1 进入caffe-windows的windows文件夹,Copy .\windows\CommonSettings.props.example to .\windows\CommonSettings.props
2 打开caffe工程,编辑CommonSettings.props文件,以下是cpu版本设置
<CpuOnlyBuild>true</CpuOnlyBuild>
<UseCuDNN>false</UseCuDNN>
<CudaVersion>7.5</CudaVersion>
<PythonSupport>false</PythonSupport>
<MatlabSupport>true</MatlabSupport>
<CudaDependencies></CudaDependencies>
<PropertyGroup Condition="'$(MatlabSupport)'=='true'">
<MatlabDir>C:\Program Files\MATLAB\R2016a</MatlabDir>
<LibraryPath>$(MatlabDir)\extern\lib\win64\microsoft;$(LibraryPath)</LibraryPath>
<IncludePath>$(MatlabDir)\extern\include;$(IncludePath)</IncludePath>
</PropertyGroup>
3 选择matcaffe项目,点击编译(会自动去下载第三方库),在Build\x64\Release会生成相应的文件
4 将上面Build\x64\Release绝对路径加入到系统环境path变量中,同时将Build\x64\Release\matcaffe加入到matlab路径中。
5 重新启动matlab,调用caffe.reset_all(),则说明ok。
>> caffe.reset_all();
Cleared 0 solvers and 0 stand-alone nets
>>
python使用
下载安装anaconda
安装protobuf:在命令行输入:pip install protobuf
使用spyder,并且设置Python路径
import caffe
caffe在matlab中使用:
function train()
solver_def_file = 'model/lenet_solver.prototxt';
caffe.set_mode_cpu();
caffe.reset_all();
solver = caffe.Solver(solver_def_file);
% solver.solve();%一次性迭代
close all;
hold on%画图用的
iter_ = solver.iter();
while iter_<10000
solver.step(1);%一步一步迭代
iter_ = solver.iter();
loss=solver.net.blobs('loss').get_data();%取训练集的loss
if iter_==1
loss_init = loss;
else
y_l=[loss_init loss];
x_l=[iter_-1, iter_];
plot(x_l, y_l, 'r-');
drawnow
loss_init = loss;
end
if mod(iter_, 100) == 0
accuracy=solver.test_nets.blobs('accuracy').get_data();%取验证集的accuracy
if iter_/100 == 1
accuracy_init = accuracy;
else
x_l=[iter_-100, iter_];
y_a=[accuracy_init accuracy];
plot(x_l, y_a,'g-');
drawnow
accuracy_init=accuracy;
end
end
end
测试
function test()
net = init_net();
im_data = 255-caffe.io.load_image('image/00082.png');
res = net.forward({im_data});
[~, idx ]= max(res{1});
disp(idx-1);
function net = init_net()
caffe.set_mode_cpu();
caffe.reset_all();
deploy = 'model/lenet_deploy.prototxt';
caffe_model = 'snapshot/lenet_iter_10000.caffemodel';
net = caffe.Net(deploy, caffe_model, 'test');
微调
function retrain()
caffe.set_mode_cpu();
caffe.reset_all();
caffe_model = 'snapshot/lenet_iter_10000.caffemodel';
solver = caffe.Solver('model/lenet_solver.prototxt');
solver.net.copy_from(caffe_model);
% solver.solve();
close all;
hold on%画图用的
iter_ = solver.iter();
while iter_<10000
solver.step(1);
iter_ = solver.iter();
loss=solver.net.blobs('loss').get_data();%取训练集的loss
if iter_==1
loss_init = loss;
else
y_l=[loss_init loss];
x_l=[iter_-1, iter_];
plot(x_l, y_l, 'r-');
drawnow
loss_init = loss;
end
if mod(iter_, 100) == 0
accuracy=solver.test_nets.blobs('accuracy').get_data();%取验证集的accuracy
if iter_/100 == 1
accuracy_init = accuracy;
else
x_l=[iter_-100, iter_];
y_a=[accuracy_init accuracy];
plot(x_l, y_a,'g-');
drawnow
accuracy_init=accuracy;
end
end
end |
|
|