org.nd4j.linalg.api.ops.impl.layers.convolution

Class SConv2D

java.lang.Object

org.nd4j.autodiff.functions.DifferentialFunction

org.nd4j.linalg.api.ops.DynamicCustomOp

org.nd4j.linalg.api.ops.impl.layers.convolution.Conv2D

org.nd4j.linalg.api.ops.impl.layers.convolution.SConv2D

All Implemented Interfaces:

CustomOp

Direct Known Subclasses:

SConv2DDerivative

public class SConv2D
extends Conv2D

Separable convolution 2D operation

Nested Class Summary

Nested classes/interfaces inherited from class org.nd4j.linalg.api.ops.DynamicCustomOp

DynamicCustomOp.DynamicCustomOpsBuilder

Field Summary

Fields inherited from class org.nd4j.linalg.api.ops.impl.layers.convolution.Conv2D

config

Fields inherited from class org.nd4j.linalg.api.ops.DynamicCustomOp

axis, bArguments, iArguments, inplaceCall, inputArguments, outputArguments, outputVariables, tArguments

Fields inherited from class org.nd4j.autodiff.functions.DifferentialFunction

dimensions, extraArgs, inPlace, sameDiff, scalarValue

Constructor Summary

Constructors

Constructor and Description
SConv2D()
SConv2D(INDArray[] inputs, INDArray[] outputs, Conv2DConfig config)
SConv2D(INDArray input, INDArray depthWeights, INDArray pointWeights, INDArray bias, INDArray output, Conv2DConfig config)
SConv2D(SameDiff sameDiff, SDVariable[] inputFunctions, Conv2DConfig conv2DConfig)

Method Summary

Modifier and Type	Method and Description
List<DataType>	calculateOutputDataTypes(List<DataType> inputDataTypes) Calculate the data types for the output arrays.
String	configFieldName() Returns the name of the field to be used for looking up field names.
List<SDVariable>	doDiff(List<SDVariable> f1) The actual implementation for automatic differentiation.
long[]	iArgs()
boolean	isConfigProperties() Returns true if the fields for this class should be looked up from a configuration class.
String	onnxName() The opName of this function in onnx
String	opName() This method returns op opName as string
String	tensorflowName() The opName of this function tensorflow
String[]	tensorflowNames() The opName of this function tensorflow

Methods inherited from class org.nd4j.linalg.api.ops.impl.layers.convolution.Conv2D

addArgs, attributeAdaptersForFunction, getValue, initConfig, initFromOnnx, initFromTensorFlow, mappingsForFunction, propertiesForFunction

Methods inherited from class org.nd4j.linalg.api.ops.DynamicCustomOp

addBArgument, addIArgument, addIArgument, addInputArgument, addOutputArgument, addTArgument, assertValidForExecution, bArgs, builder, calculateOutputShape, getBArgument, getDescriptor, getIArgument, getInputArgument, getOutputArgument, getTArgument, inputArguments, numBArguments, numIArguments, numInputArguments, numOutputArguments, numTArguments, opHash, opNum, opType, outputArguments, outputVariables, outputVariables, removeIArgument, removeInputArgument, removeOutputArgument, removeTArgument, setInputArgument, setInputArguments, setOutputArgument, tArgs, toString, wrapFilterNull, wrapOrNull

Methods inherited from class org.nd4j.autodiff.functions.DifferentialFunction

arg, arg, argNames, args, diff, dup, equals, f, getNumOutputs, hashCode, larg, onnxNames, outputVariable, outputVariablesNames, rarg, replaceArg, resolvePropertiesFromSameDiffBeforeExecution, setInstanceId, setPropertiesForFunction, setValueFor

Methods inherited from class java.lang.Object

clone, finalize, getClass, notify, notifyAll, wait, wait, wait

Methods inherited from interface org.nd4j.linalg.api.ops.CustomOp

isInplaceCall

Constructor Detail

SConv2D

public SConv2D(SameDiff sameDiff,
               SDVariable[] inputFunctions,
               Conv2DConfig conv2DConfig)

SConv2D

public SConv2D(INDArray[] inputs,
               INDArray[] outputs,
               Conv2DConfig config)

SConv2D

public SConv2D(@NonNull
               INDArray input,
               @NonNull
               INDArray depthWeights,
               INDArray pointWeights,
               INDArray bias,
               INDArray output,
               @NonNull
               Conv2DConfig config)

SConv2D

public SConv2D()

Method Detail

opName

public String opName()

Description copied from class: DynamicCustomOp

This method returns op opName as string

Specified by:

opName in interface CustomOp

Overrides:

opName in class Conv2D

Returns:

doDiff

public List<SDVariable> doDiff(List<SDVariable> f1)

Description copied from class: DifferentialFunction

The actual implementation for automatic differentiation.

Overrides:

doDiff in class Conv2D

Returns:

iArgs

public long[] iArgs()

Specified by:

iArgs in interface CustomOp

Overrides:

iArgs in class Conv2D

isConfigProperties

public boolean isConfigProperties()

Description copied from class: DifferentialFunction

Returns true if the fields for this class should be looked up from a configuration class.

Overrides:

isConfigProperties in class Conv2D

Returns:

configFieldName

public String configFieldName()

Description copied from class: DifferentialFunction

Returns the name of the field to be used for looking up field names. This should be used in conjunction with DifferentialFunction.isConfigProperties() to facilitate mapping fields for model import.

Overrides:

configFieldName in class Conv2D

Returns:

tensorflowNames

public String[] tensorflowNames()

Description copied from class: DifferentialFunction

The opName of this function tensorflow

Overrides:

tensorflowNames in class Conv2D

Returns:

onnxName

public String onnxName()

Description copied from class: DifferentialFunction

The opName of this function in onnx

Overrides:

onnxName in class Conv2D

Returns:

tensorflowName

public String tensorflowName()

Description copied from class: DifferentialFunction

The opName of this function tensorflow

Overrides:

tensorflowName in class Conv2D

Returns:

calculateOutputDataTypes

public List<DataType> calculateOutputDataTypes(List<DataType> inputDataTypes)

Description copied from class: DifferentialFunction

Calculate the data types for the output arrays. Though datatypes can also be inferred from DifferentialFunction.calculateOutputShape(), this method differs in that it does not require the input arrays to be populated. This is important as it allows us to do greedy datatype inference for the entire net - even if arrays are not available.

Overrides:

calculateOutputDataTypes in class Conv2D

Parameters:

inputDataTypes - The data types of the inputs

Returns:

The data types of the outputs