Conv3DTranspose

class Conv3DTranspose(filters: Int, kernelSize: IntArray, strides: IntArray, dilations: IntArray, activation: Activations, kernelInitializer: Initializer, biasInitializer: Initializer, kernelRegularizer: Regularizer?, biasRegularizer: Regularizer?, activityRegularizer: Regularizer?, padding: ConvPadding, useBias: Boolean, name: String) : ConvTranspose, NoGradients

3D convolution transpose layer.

This is an operation going in the opposite direction of a normal convolution: it transforms a tensor shaped like an output of some convolution into tensor that has the shape of the input.

This layer expects input data of size (N, D, H, W, C) where

N - batch size
D - depth
H - height
W - width
C - number of channels

Note: providing explicit output padding is currently not supported. Dilation values greater than 1 are not supported on cpu.

Since

0.4

Parameters

filters

dimensionality of the output space (i.e. the number of filters in the convolution)

activation

activation function

kernelInitializer

initializer for the kernel

biasInitializer

initializer for the bias

kernelRegularizer

regularizer for the kernel

biasRegularizer

regularizer for the bias

activityRegularizer

regularizer function applied to the output of the layer

padding

type of padding to use

useBias

a flag that specifies if the bias should be used

name

custom layer name

Constructors

Conv3DTranspose

fun Conv3DTranspose(filters: Int = 3, kernelSize: Int = 3, strides: Int = 1, dilations: Int = 1, activation: Activations = Activations.Relu, kernelInitializer: Initializer = HeNormal(), biasInitializer: Initializer = HeUniform(), kernelRegularizer: Regularizer? = null, biasRegularizer: Regularizer? = null, activityRegularizer: Regularizer? = null, padding: ConvPadding = ConvPadding.SAME, useBias: Boolean = true, name: String = "")

Conv3DTranspose

fun Conv3DTranspose(filters: Int = 3, kernelSize: IntArray = intArrayOf(3, 3, 3), strides: IntArray = intArrayOf(1, 1, 1, 1, 1), dilations: IntArray = intArrayOf(1, 1, 1, 1, 1), activation: Activations = Activations.Relu, kernelInitializer: Initializer = HeNormal(), biasInitializer: Initializer = HeUniform(), kernelRegularizer: Regularizer? = null, biasRegularizer: Regularizer? = null, activityRegularizer: Regularizer? = null, padding: ConvPadding = ConvPadding.SAME, useBias: Boolean = true, name: String = "")

Functions

build

open override fun build(tf: Ops, input: Operand<Float>, isTraining: Operand<Boolean>, numberOfLosses: Operand<Float>?): Operand<Float>

open fun build(tf: Ops, input: List<Operand<Float>>, isTraining: Operand<Boolean>, numberOfLosses: Operand<Float>?): Operand<Float>

Extend this function to define variables in the layer and compute layer output.

invoke

operator fun invoke(vararg layers: Layer): Layer

Important part of functional API. It takes layers as input and saves them to the inboundLayers of the given layer.

toString

open override fun toString(): String

Properties

activation

open override val activation: Activations

activityRegularizer

open override val activityRegularizer: Regularizer? = null

biasInitializer

open override val biasInitializer: Initializer

biasRegularizer

open override val biasRegularizer: Regularizer? = null

dilations

open override val dilations: IntArray

dilations of the convolution for each dimension of the input tensor (five numbers). Currently, dilation values greater than 1 are not supported on cpu.

dimensions

val dimensions: Int

dimensionality of this convolution operation

filters

open override val filters: Int = 3

hasActivation

open override val hasActivation: Boolean

Returns True, if layer has internal activation function.

inboundLayers

var inboundLayers: MutableList<Layer>

Returns inbound layers.

kernelInitializer