caffe的mnist数据集测试

准备

1. 前文训练得到caffemodel
   
2. 均值文件mean.binaryproto

步骤

1. 新建mnist_mean.bat

D:\Study\Caffe\caffe-windows\Build\x64\Release\compute_image_mean.exe 
D:\Study\Caffe\caffe-windows\examples\mnist\mnist_train_lmdb mean.binaryproto 
--backend=lmdb
pause

运行后会在此bat文件存放的目录下生成一个mean.binaryproto

2. 新建mnist_test.bat

D:\Study\Caffe\caffe-windows\Build\x64\Release\caffe.exe
test
--model=D:\Study\Caffe\caffe-windows\examples\mnist\lenet_train_test.prototxt 
-weights=D:\Study\Caffe\caffe-windows\examples\mnist\lenet_iter_10000.caffemodel
pause

3. 修改lenet_train_test.prototxt

4. 将mnist_test.bat和mean.binaryproto放到目录D:\Study\Caffe\caffe-windows\examples\mnist下，运行mnist_test.bat