无法在Keras中复制matconvnet CNN架构
[英]Can't replicate a matconvnet CNN architecture in Keras
问题描述
I have the following architecture of a Convolutional Neural Network in matconvnet which I use to train on my own data: 
我在matconvnet中有以下卷积神经网络的架构，我用它来训练我自己的数据：
function net = cnn_mnist_init(varargin)
% CNN_MNIST_LENET Initialize a CNN similar for MNIST
opts.batchNormalization = false ;
opts.networkType = 'simplenn' ;
opts = vl_argparse(opts, varargin) ;
f= 0.0125 ;
net.layers = {} ;
net.layers{end+1} = struct('name','conv1',...
'type', 'conv', ...
'weights', {{f*randn(3,3,1,64, 'single'), zeros(1, 64, 'single')}}, ...
'stride', 1, ...
'pad', 0,...
'learningRate', [1 2]) ;
net.layers{end+1} = struct('name','pool1',...
'type', 'pool', ...
'method', 'max', ...
'pool', [3 3], ...
'stride', 1, ...
'pad', 0);
net.layers{end+1} = struct('name','conv2',...
'type', 'conv', ...
'weights', {{f*randn(5,5,64,128, 'single'),zeros(1,128,'single')}}, ...
'stride', 1, ...
'pad', 0,...
'learningRate', [1 2]) ;
net.layers{end+1} = struct('name','pool2',...
'type', 'pool', ...
'method', 'max', ...
'pool', [2 2], ...
'stride', 2, ...
'pad', 0) ;
net.layers{end+1} = struct('name','conv3',...
'type', 'conv', ...
'weights', {{f*randn(3,3,128,256, 'single'),zeros(1,256,'single')}}, ...
'stride', 1, ...
'pad', 0,...
'learningRate', [1 2]) ;
net.layers{end+1} = struct('name','pool3',...
'type', 'pool', ...
'method', 'max', ...
'pool', [3 3], ...
'stride', 1, ...
'pad', 0) ;
net.layers{end+1} = struct('name','conv4',...
'type', 'conv', ...
'weights', {{f*randn(5,5,256,512, 'single'),zeros(1,512,'single')}}, ...
'stride', 1, ...
'pad', 0,...
'learningRate', [1 2]) ;
net.layers{end+1} = struct('name','pool4',...
'type', 'pool', ...
'method', 'max', ...
'pool', [2 2], ...
'stride', 1, ...
'pad', 0) ;
net.layers{end+1} = struct('name','ip1',...
'type', 'conv', ...
'weights', {{f*randn(1,1,256,256, 'single'), zeros(1,256,'single')}}, ...
'stride', 1, ...
'pad', 0,...
'learningRate', [1 2]) ;
net.layers{end+1} = struct('name','relu',...
'type', 'relu');
net.layers{end+1} = struct('name','classifier',...
'type', 'conv', ...
'weights', {{f*randn(1,1,256,2, 'single'), zeros(1,2,'single')}}, ...
'stride', 1, ...
'pad', 0,...
'learningRate', [1 2]) ;
net.layers{end+1} = struct('name','loss',...
'type', 'softmaxloss') ;
% optionally switch to batch normalization
if opts.batchNormalization
net = insertBnorm(net, 1) ;
net = insertBnorm(net, 4) ;
net = insertBnorm(net, 7) ;
net = insertBnorm(net, 10) ;
net = insertBnorm(net, 13) ;
end
% Meta parameters
net.meta.inputSize = [28 28 1] ;
net.meta.trainOpts.learningRate = [0.01*ones(1,10) 0.001*ones(1,10) 0.0001*ones(1,10)];
disp(net.meta.trainOpts.learningRate);
pause;
net.meta.trainOpts.numEpochs = length(net.meta.trainOpts.learningRate) ;
net.meta.trainOpts.batchSize = 256 ;
net.meta.trainOpts.momentum = 0.9 ;
net.meta.trainOpts.weightDecay = 0.0005 ;
% --------------------------------------------------------------------
function net = insertBnorm(net, l)
% --------------------------------------------------------------------
assert(isfield(net.layers{l}, 'weights'));
ndim = size(net.layers{l}.weights{1}, 4);
layer = struct('type', 'bnorm', ...
'weights', {{ones(ndim, 1, 'single'), zeros(ndim, 1, 'single')}}, ...
'learningRate', [1 1], ...
'weightDecay', [0 0]) ;
net.layers{l}.biases = [] ;
net.layers = horzcat(net.layers(1:l), layer, net.layers(l+1:end)) ;
What I want to do is building the same architecture in Keras, that is what I tried so far: 我想要做的是在Keras建立相同的架构，这是我到目前为止所尝试的：
model = Sequential()
model.add(Conv2D(64, (3, 3), strides=1, input_shape=input_shape))
model.add(MaxPooling2D(pool_size=(3, 3), strides=1))
model.add(Conv2D(128, (5, 5), strides=1))
model.add(MaxPooling2D(pool_size=(2, 2), strides=2))
model.add(Conv2D(256, (3, 3), strides=1))
model.add(MaxPooling2D(pool_size=(3, 3), strides=1))
model.add(Conv2D(512, (5, 5), strides=1))
model.add(MaxPooling2D(pool_size=(2, 2), strides=1))
model.add(Conv2D(256, (1, 1)))
convout1=Activation('relu')
model.add(convout1)
model.add(Flatten())
model.add(Dense(num_classes, activation='softmax'))
opt = keras.optimizers.rmsprop(lr=0.0001, decay=0.0005)
model.compile(loss='categorical_crossentropy', optimizer=opt, metrics=['binary_accuracy'])
However, when I run the matconvnet network I have 87% accuracy and if I run the keras version I have 77% accuracy. 但是，当我运行matconvnet网络时，我有87％的准确率，如果我运行keras版本，我有77％的准确率。 If they are supposed to be the same network and the data is the same, where is the difference? 如果他们应该是同一个网络并且数据是相同的，那么区别在哪里？ What is wrong in my Keras architecture? 我的Keras架构出了什么问题？
1 个解决方案
解决方案1
1 已采纳 2017-07-03 22:14:07
In your MatConvNet version, you use SGD with momentum. 在MatConvNet版本中，您可以使用SGD。
In Keras, you use rmsprop 在Keras中，您使用rmsprop
With a different learning rule you should try different learning rates. 根据不同的学习规则，您应该尝试不同的学习率。 Also sometimes momentum is helpful when training a CNN. 在培训CNN时，有时候动量也很有帮助。
Could you try the SGD+momentum in Keras and let me know what happens? 你能尝试一下Keras的SGD +势头，让我知道会发生什么吗？
Another thing that might be different is that the initialization. 另一件可能不同的是初始化。 for example in MatConvNet you use gaussian initialization with f= 0.0125 as the standard deviation. 例如，在MatConvNet中，您使用高斯初始化，其中f = 0.0125作为标准偏差。 In Keras I'm not sure about the default initialization. 在Keras我不确定默认初始化。
In general, if you don't use batch normalization, the network is prone to many numerical issues. 通常，如果不使用批量规范化，则网络容易出现许多数值问题。 If you use batch normalization in both networks, I bet the results would be similar. 如果你在两个网络中使用批量标准化，我敢打赌结果会是相似的。 Is there any reason you don't want to use batch normalization? 你有什么理由不想使用批量标准化吗？
声明:本站的技术帖子网页，遵循CC BY-SA 4.0协议，如果您需要转载，请注明本站网址或者原文地址。任何问题请咨询:yoyou2525@163.com.