public override AbstractLayer Compile(AbstractLayer previousLayer)
{
if (previousLayer == null && this.inputDimension == null)
{
string msg = "Dimension of first layer of network not set!";
Utils.ThrowException(msg);
}
else if (inputDimension == null)
{
inputDimension = previousLayer.outputDimension;
index = 0;
}
else if (previousLayer != null)
{
inputDimension = previousLayer.outputDimension;
index = previousLayer.Index + 1;
}
TestHyperParametersAndSetOutputSize(inputDimension.width);
outputDimension = new Dimension(inputDimension.imageCount, inputDimension.depth,
outputSize, outputSize);
compiled = true;
return(this);
}
public override AbstractLayer Compile(AbstractLayer previousLayer)
{
if (previousLayer == null && inputDimension == null)
{
string msg = "Dimension of first layer of network not set!";
Utils.ThrowException(msg);
}
else if (previousLayer != null)
{
inputDimension = previousLayer.outputDimension;
index = previousLayer.Index + 1;
}
TestHyperParametersAndSetOutputSize(this.inputDimension.width,
this.inputDimension.height, this.inputDimension.depth);
outputDimension = new Dimension(this.inputDimension.imageCount, filterCount,
1, outputSize);
#region Init initializers
if (biasInitializer == null)
{
biasInitializer = new ConstantBiasInitializer(0d);
}
biases = biasInitializer.Initialize(filterCount);
// https://stats.stackexchange.com/questions/198840/cnn-xavier-weight-initialization
// https://github.com/keras-team/keras/blob/998efc04eefa0c14057c1fa87cab71df5b24bf7e/keras/initializations.py
int fanIn = filterSize * filterSize * inputDimension.depth;
int fanOut = filterCount;
if (weightInitializer == null)
{
weightInitializer = new XavierWeightInitializer(fanIn, fanOut);
}
#endregion Init initializers
filters = weightInitializer.Initialize(new Dimension(filterCount,
inputDimension.depth, inputDimension.width, filterSize));
// activations will be init
activations = Utils.Init4dArr(inputDimension.imageCount,
filterCount, inputDimension.width, outputSize);
compiled = true;
return(this);
}
public override AbstractLayer Compile(AbstractLayer previousLayer)
{
if (previousLayer == null && inputDimension == null)
{
string msg = "Dimension of first layer of network not set!";
Utils.ThrowException(msg);
}
else if (previousLayer != null)
{
inputDimension = previousLayer.outputDimension;
index = previousLayer.Index + 1;
}
outputDimension = inputDimension;
compiled = true;
return(this);
}
public override AbstractLayer Compile(AbstractLayer previousLayer)
{
if (previousLayer == null)
{
string msg = "Dropout layer cannot be first layer!";
Utils.ThrowException(msg);
}
if (previousLayer.outputDimension.depth != 1 ||
previousLayer.outputDimension.width != 1)
{
string msg = "Dropout layer is adapted only for flatten layers!";
Utils.ThrowException(msg);
}
index = previousLayer.Index + 1;
compiled = true;
outputDimension = previousLayer.outputDimension;
return(this);
}
public override AbstractLayer Compile(AbstractLayer previousLayer)
{
if (previousLayer == null && this.inputDimension == null)
{
string msg = "Dimension of first layer of network not set!";
Utils.ThrowException(msg);
}
else if (previousLayer != null)
{
inputDimension = previousLayer.outputDimension;
index = previousLayer.Index + 1;
}
size = inputDimension.depth * inputDimension.height * inputDimension.width;
outputDimension = new Dimension(inputDimension.imageCount, 1, 1, size);
compiled = true;
return(this);
}
public override AbstractLayer Compile(AbstractLayer previousLayer)
{
if (previousLayer == null)
{
string msg = "Dimension of first layer of network not set!";
Utils.ThrowException(msg);
}
else if (previousLayer != null)
{
inputDimension = previousLayer.outputDimension;
index = previousLayer.Index + 1;
}
inputDimension = previousLayer.outputDimension;
outputDimension = new Dimension(inputDimension.imageCount, 1, 1, numNeurons);
activations = Utils.Init4dArr(inputDimension.imageCount, 1, 1, numNeurons);
#region Init initializers
if (biasInitializer == null)
{
biasInitializer = new ConstantBiasInitializer(0d);
}
biases = biasInitializer.Initialize(numNeurons);
int fanIn = inputDimension.height;
int fanOut = numNeurons;
if (weightInitializer == null)
{
weightInitializer = new XavierWeightInitializer(fanIn, fanOut);
}
weights = weightInitializer.Initialize(new Dimension(inputDimension.height, 1, 1, numNeurons));
#endregion Init initializers
compiled = true;
return(this);
}
/// <summary> /// Method which check the dimension of the layer and initializes /// all possible values in the layer - e.g. weights and biases. /// </summary> /// <param name="previousLayer">previous layer in the model</param> /// <returns>current layer</returns> public abstract AbstractLayer Compile(AbstractLayer previousLayer);
