dlpy.layers.Segmentation

class dlpy.layers.Segmentation(name=None, act=None, error=None, target_scale=1.0, src_layers=None, output_image_type=None, output_image_prob=None, **kwargs)

Bases: dlpy.layers.Layer

Segmentation layer

Parameters:
name : string, optional

Specifies the name of the layer.

act : string, optional

Specifies the activation function. possible values: [ AUTO, IDENTITY, LOGISTIC, SIGMOID, TANH, RECTIFIER, RELU, SOFPLUS, ELU, LEAKY, FCMP ]
Default: AUTO

fcmp_act : string, optional

Specifies the FCMP activation function for the layer.

src_layers : iterable Layer, optional

Specifies the layers directed to this layer.

height : int, required

Specifies the height of the input data. By default the height is determined automatically when the model training begins.

width : int, required

Specifies the width of the input data. By default the width is determined automatically when the model training begins.

depth : int, required

Specifies the depth of the feature maps.

target_scale : double

Specifies the factor used to scale target values for a segmentation layer.

src_layers : iterable Layer, optional

Specifies the layers directed to this layer.

output_image_type: string, optional

Specifies the output image type of this layer. possible values: [ WIDE, PNG, BASE64 ]
Default: WIDE

output_image_prob: bool, options

Does not include probabilities if doing classification (default).

Returns:
Segmentation
__init__(name=None, act=None, error=None, target_scale=1.0, src_layers=None, output_image_type=None, output_image_prob=None, **kwargs)

Initialize self. See help(type(self)) for accurate signature.

Methods

__init__([name, act, error, target_scale, …]) Initialize self.
count_instances()
format_name([block_num, local_count]) Format the name of the layer
get_number_of_instances()
to_model_params() Convert the model configuration to CAS action parameters