DepthwiseConv2D

Public technical reference documentation from Graiphic. For current SOTA installation and toolkit guides, visit the documentation center.

Description

Adds the weights of the DepthwiseConv2D layer to the weights table. Type : polymorphic.

Documentation illustration (illustration unavailable in the archive).

Input parameters

Weights in : array

name : string, name of layer.
Documentation illustration (illustration unavailable in the archive). weights : variant, weights values.

Documentation illustration (illustration unavailable in the archive).

name : string, name of layer.
filters_depthwise : array, 4D values. filters_depthwise = [channels, 1, size[0], size[1]].
biases : array, 1D values. biases = [channels].

Output parameters

Weights out : array

name : string, name of layer.
Documentation illustration (illustration unavailable in the archive). weights : variant, weights values.

Documentation illustration (illustration unavailable in the archive).

Dimension

filters_depthwise = [channels, 1, size[0], size[1]]

The size of filters_depthwise depends on the input of the DepthwiseConv2D layer and the parameters size.
For example if the input of the layer has a size of [batch_size = 10, channels = 8, rows = 5, cols = 5] and size the value [3, 3] then filters will have a size of [channels = 8, 1, size[0] = 3, size[1] = 3].

biases = [channels]

The size of biases depends on the parameter size of the DepthwiseConv2D layer.
For example, if the input of the layer has a size of [batch_size = 10, channels = 8, rows = 5, cols = 5] then biases will have a size of [channels = 8].

Example

All these exemples are snippets PNG, you can drop these Snippet onto the block diagram and get the depicted code added to your VI (Do not forget to install Deep Learning library to run it).