CN03 Depths in Convolutional Neural Network ( ConvNet / CNN Architecture )
Depths in Convolutional Neural Networks ( ConvNets / CNNs )

The term depth is used in three different contexts in ConvNets.

Depth of Network: First one and the most common is in the context of depth of network as in Deep Network or Deep Learning. In this context it means number of layers in the ConvNet. More the number of layers higher the depth or Deeper the network.

Depth of a Layer: Second is Depth of a particular layer. The way neurons are arranged in a layer, each individual layer also has depth dimension in it. The neurons in each layer are arranged in three dimensional array. Width x Height x Depth. These dimensions can be quantified by number of neurons in each dimension. This is the reason why layers in ConvNets are often referred to as volumes, i.e., input layer is also called input volume and output layer is also called output volume. A hidden layer can act both as an input as well as output layer so depending on its role it may be called input volume or output volume.

Depth of Filter: The weights of links between neurons of two layers are stored in a filter or a kernel. A filter or kernel in general in nothing but a multidimensional array of numbers. In case of ConvNets a filter can be a 3 dimensional array. Width x Height x Depth. Width and height of a filter are much smaller than the width and height of input volume but depth of filter is kept same as depth of input volume.

What is depth of network in CNN ?
What is meaning of deep network ?
What is deep learning ?
What is a tensor ?
What is a tensor in CNN ?
Architecture of CNN
What is the architecture of layer in CNN ?
Why nodes in CNN are arranged in three dimensional array ?
What is depth of layer in CNN ?
What is receptive field in CNN ?
What is a filter in CNN ?
What is a kernel in CNN ?
Why do we use filter in CNN ?
What is the purpose of using filter in CNN ?
Why do we use kernel in CNN ?
What is the purpose of using kernel in CNN ?
What is depth of network in ConvNet ?
What is meaning of deep network ?
What is deep learning ?
What is a tensor ?
What is a tensor in ConvNet ?
Architecture of ConvNet
What is the architecture of layer in ConvNet ?
Why nodes in ConvNet are arranged in three dimensional array ?
What is depth of layer in ConvNet ?
What is receptive field in ConvNet ?
What is a filter in ConvNet ?
What is a kernel in ConvNet ?
Why do we use filter in ConvNet ?
What is the purpose of using filter in ConvNet ?
Why do we use kernel in ConvNet ?
What is the purpose of using kernel in ConvNet ?
What is depth of network in Convolutional Neural Network ?
What is meaning of deep network ?
What is deep learning ?
What is a tensor ?
What is a tensor in Convolutional Neural Network ?
Architecture of Convolutional Neural Network
What is the architecture of layer in Convolutional Neural Network ?
Why nodes in Convolutional Neural Network are arranged in three dimensional array ?
What is depth of layer in Convolutional Neural Network ?
What is receptive field in Convolutional Neural Network ?
What is a filter in Convolutional Neural Network ?
What is a kernel in Convolutional Neural Network ?
Why do we use filter in Convolutional Neural Network ?
What is the purpose of using filter in Convolutional Neural Network ?
Why do we use kernel in Convolutional Neural Network ?
What is the purpose of using kernel in Convolutional Neural Network ?

свернуть