Operator Reference
create_dl_layer_concat (Operator)
create_dl_layer_concat — Create a concatenation layer.
Signature
create_dl_layer_concat( : : DLLayerInputs, LayerName, Axis, GenParamName, GenParamValue : DLLayerConcat)
Description
The operator create_dl_layer_concat creates a concatenation layer
whose handle is returned in DLLayerConcat.
The parameter DLLayerInputs determines the feeding input layers.
This layer expects multiple layers as input.
The parameter LayerName sets an individual layer name.
Note that if creating a model using create_dl_model each layer of
the created network must have a unique name.
A concatenation layer concatenates the data tensors of the input layers
in DLLayerInputs and returns a single data tensor
DLLayerConcat.
The parameter Axis specifies along which dimension the
inputs should be concatenated. The supported options for Axis
are:
- 'batch':
-
Concatenation is applied along the
batch-dimension.Example: if you concatenate two inputs A and B of shape (h, w, d, b) = (1, 1, 1, 2), where A = [A0, A1] and B = [B0, B1], you obtain the output [A0, A1, B0, B1] with shape (1, 1, 1, 4).
- 'batch_interleaved':
-
Concatenation is applied along the
depth-dimension, but the output is reshaped as if the data was concatenated along thebatch-dimension. For this dimension, all inputs need to have exactly the same shape.Note that when the input
batch_sizeis 1, the concatenation is identical for 'batch' and 'batch_interleaved'.Example: if you concatenate two inputs A and B of shape (h, w, d, b) = (1, 1, 1, 2), where A = [A0, A1] and B = [B0, B1], you obtain the output [A0, B0, A1, B1] with shape (1, 1, 1, 4).
- 'depth':
-
Concatenation is applied along the
depth-dimension.Example: if you concatenate two inputs A and B of shape (h, w, d, b) = (1, 1, 1, 2), where A = [A0, A1] and B = [B0, B1], you obtain the output [A0, A1, B0, B1] with shape (1, 1, 2, 2).
- 'height':
Concatenation is applied along the
height-dimension.- 'width':
Concatenation is applied along the
width-dimension.
Note that all non-concatenated dimensions must be equal for all input data tensors.
The following generic parameters GenParamName and the corresponding
values GenParamValue are supported:
- 'is_inference_output':
-
Determines whether
apply_dl_modelwill include the output of this layer in the dictionaryDLResultBatcheven without specifying this layer inOutputs('true') or not ('false').Default: 'false'
Certain parameters of layers created using this operator
create_dl_layer_concat can be set and retrieved using
further operators.
The following tables give an overview, which parameters can be set
using set_dl_model_layer_param and which ones can be retrieved
using get_dl_model_layer_param or get_dl_layer_param.
Note, the operators set_dl_model_layer_param and
get_dl_model_layer_param require a model created by
create_dl_model.
| Layer Parameters | set |
get |
|---|---|---|
'input_layer' (DLLayerInputs) |
x
|
|
'name' (LayerName) |
x |
x
|
'output_layer' (DLLayerConcat) |
x
|
|
| 'shape' | x
|
|
| 'type' | x
|
| Generic Layer Parameters | set |
get |
|---|---|---|
| 'is_inference_output' | x |
x
|
| 'num_trainable_params' | x
|
Execution Information
- Multithreading type: reentrant (runs in parallel with non-exclusive operators).
- Multithreading scope: global (may be called from any thread).
- Processed without parallelization.
Parameters
DLLayerInputs (input_control) dl_layer(-array) → (handle)
Feeding input layers.
LayerName (input_control) string → (string)
Name of the output layer.
Axis (input_control) string → (string)
Dimension along which the input layers are concatenated.
Default: 'depth'
List of values: 'batch', 'batch_interleaved', 'depth', 'height', 'width'
GenParamName (input_control) attribute.name(-array) → (string)
Generic input parameter names.
Default: []
List of values: 'is_inference_output'
GenParamValue (input_control) attribute.value(-array) → (string / integer / real)
Generic input parameter values.
Default: []
Suggested values: 'true', 'false'
DLLayerConcat (output_control) dl_layer → (handle)
Concatenation layer.
Module
Deep Learning Professional