dlhdl.layer.splitLayer

Split layer

Since R2025a

Description

Use dlhdl.layer.splitLayer objects to create a split layer. You can use split layers to divide a layer output into equal parts along the channel dimension of the image.

Creation

Syntax

layer = dlhdl.layer.splitLayer(Name,numOutputGroups)

Description

layer = dlhdl.layer.splitLayer(Name,numOutputGroups) creates a split layer, sets the Name property, and splits the output of the layer into equal parts specified by the numOutputGroups argument along the channel dimension of the image.

Input Arguments

numOutputGroups — Number of groups to split the layer output into

2 (default) | positive integer

Number of groups to split layer output into, specified as a positive integer.

Example: 2

Properties

Layer

Name — Layer name

'' (default) | character vector | string scalar

Layer name, specified as a character vector or a string scalar. For Layer array input, the trainnet and dlnetwork functions automatically assign names to unnamed layers.

The splitLayer object stores this property as a character vector.

Data Types: char | string

NumInputs — Number of inputs

Read-only: 1 (default)

This property is read-only.

Number of inputs to the layer, stored as 1. This layer accepts a single input only.

Data Types: double

InputNames — Input names

Read-only: {'in'} (default)

This property is read-only.

Input names, stored as {'in'}. This layer accepts a single input only.

Data Types: cell

NumOutputs — Number of outputs

Read-only: 2 (default) | positive integer

This property is read-only.

Number of outputs from the layer, returned as positive integer. Specify the number of outputs using numOutputGroups.

OutputNames — Output names

Read-only: string array | cell array of character vectors

This property is read-only.

Output names of the layer, returned as a string array or a cell array of character vectors.

If numOutputGroups is 1, then the software sets OutputNames to {'out'}. If numOutputGroups is greater than 1, then the software sets OutputNames to {'out1',...,'outM'}, where M is the number of outputs.

Examples

Create Split Layer

Create a split layer with the name split.

s = dlhdl.layer.splitLayer("split",2)

s = 

  splitLayer with properties:

           Name: 'split'
     NumOutputs: 2
    OutputNames: {'out1'  'out2'}

   Learnable Parameters
    No properties.

   State Parameters
    No properties.

  Show all properties

Include the split layer in a Layer array.

 imageSize = [20 20 16];
layers = [
                imageInputLayer(imageSize,"Name","imageinput","Normalization","none")
                convolution2dLayer(3,16,"Padding","same")
                dlhdl.layer.splitLayer("split",2)
                maxPooling2dLayer(3,'Padding','same','Name','pool')]

layers = 

  4×1 Layer array with layers:

     1   'imageinput'   Image Input                20×20×16 images
     2   ''             2-D Convolution            16 3×3 convolutions with stride [1  1] and padding 'same'
     3   'split'        dnnfpga.layer.splitLayer   dnnfpga.layer.splitLayer
     4   'pool'         2-D Max Pooling            3×3 max pooling with stride [1  1] and padding 'same'

Version History

Introduced in R2025a