org.deeplearning4j.nn.conf.layers

Class InputTypeUtil

java.lang.Object

org.deeplearning4j.nn.conf.layers.InputTypeUtil

public class InputTypeUtil
extends Object

Method Summary

Modifier and Type	Method and Description
static void	convertMultipleTypes(InputType[] vertexInputs) Convert multiple types when multiple are found.
static InputType	getOutputTypeCnn1DLayers(InputType inputType, int kH, int sH, int padH, int dilation, ConvolutionMode convolutionMode, long outputDepth, long layerIdx, String layerName, Class<?> layerClass)
static InputType	getOutputTypeCnn3DLayers(InputType inputType, Convolution3D.DataFormat dataFormat, int[] kernelSize, int[] stride, int[] padding, int[] dilation, ConvolutionMode convolutionMode, long outputChannels, long layerIdx, String layerName, Class<?> layerClass)
static InputType	getOutputTypeCnnLayers(InputType inputType, int[] kernelSize, int[] stride, int[] padding, int[] dilation, ConvolutionMode convolutionMode, long outputDepth, long layerIdx, String layerName, Class<?> layerClass) Deprecated. Use getOutputTypeCnnLayers(InputType, int[], int[], int[], int[], ConvolutionMode, long, long, String, CNN2DFormat, Class)
static InputType	getOutputTypeCnnLayers(InputType inputType, int[] kernelSize, int[] stride, int[] padding, int[] dilation, ConvolutionMode convolutionMode, long outputDepth, long layerIdx, String layerName, CNN2DFormat format, Class<?> layerClass)
static InputType	getOutputTypeDeconv3dLayer(InputType inputType, int[] kernelSize, int[] stride, int[] padding, int[] dilation, ConvolutionMode convolutionMode, Convolution3D.DataFormat dataFormat, long outputDepth, long layerIdx, String layerName, Class<?> layerClass)
static InputType	getOutputTypeDeconvLayer(InputType inputType, int[] kernelSize, int[] stride, int[] padding, int[] dilation, ConvolutionMode convolutionMode, long outputDepth, long layerIdx, String layerName, Class<?> layerClass)
static InputPreProcessor	getPreProcessorForInputTypeCnn3DLayers(InputType inputType, String layerName) Utility method for determining the appropriate preprocessor for CNN layers, such as ConvolutionLayer and SubsamplingLayer
static InputPreProcessor	getPreProcessorForInputTypeCnnLayers(InputType inputType, String layerName) Utility method for determining the appropriate preprocessor for CNN layers, such as ConvolutionLayer and SubsamplingLayer
static InputPreProcessor	getPreprocessorForInputTypeRnnLayers(InputType inputType, RNNFormat rnnDataFormat, String layerName)

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Method Detail

getOutputTypeDeconvLayer

public static InputType getOutputTypeDeconvLayer(InputType inputType,
                                                 int[] kernelSize,
                                                 int[] stride,
                                                 int[] padding,
                                                 int[] dilation,
                                                 ConvolutionMode convolutionMode,
                                                 long outputDepth,
                                                 long layerIdx,
                                                 String layerName,
                                                 Class<?> layerClass)

getOutputTypeDeconv3dLayer

public static InputType getOutputTypeDeconv3dLayer(InputType inputType,
                                                   int[] kernelSize,
                                                   int[] stride,
                                                   int[] padding,
                                                   int[] dilation,
                                                   ConvolutionMode convolutionMode,
                                                   Convolution3D.DataFormat dataFormat,
                                                   long outputDepth,
                                                   long layerIdx,
                                                   String layerName,
                                                   Class<?> layerClass)

getOutputTypeCnn3DLayers

public static InputType getOutputTypeCnn3DLayers(InputType inputType,
                                                 Convolution3D.DataFormat dataFormat,
                                                 int[] kernelSize,
                                                 int[] stride,
                                                 int[] padding,
                                                 int[] dilation,
                                                 ConvolutionMode convolutionMode,
                                                 long outputChannels,
                                                 long layerIdx,
                                                 String layerName,
                                                 Class<?> layerClass)

getOutputTypeCnn1DLayers

public static InputType getOutputTypeCnn1DLayers(InputType inputType,
                                                 int kH,
                                                 int sH,
                                                 int padH,
                                                 int dilation,
                                                 ConvolutionMode convolutionMode,
                                                 long outputDepth,
                                                 long layerIdx,
                                                 String layerName,
                                                 Class<?> layerClass)

getOutputTypeCnnLayers

@Deprecated
public static InputType getOutputTypeCnnLayers(InputType inputType,
                                                            int[] kernelSize,
                                                            int[] stride,
                                                            int[] padding,
                                                            int[] dilation,
                                                            ConvolutionMode convolutionMode,
                                                            long outputDepth,
                                                            long layerIdx,
                                                            String layerName,
                                                            Class<?> layerClass)

Deprecated. Use getOutputTypeCnnLayers(InputType, int[], int[], int[], int[], ConvolutionMode, long, long, String, CNN2DFormat, Class)

getOutputTypeCnnLayers

public static InputType getOutputTypeCnnLayers(InputType inputType,
                                               int[] kernelSize,
                                               int[] stride,
                                               int[] padding,
                                               int[] dilation,
                                               ConvolutionMode convolutionMode,
                                               long outputDepth,
                                               long layerIdx,
                                               String layerName,
                                               CNN2DFormat format,
                                               Class<?> layerClass)

getPreProcessorForInputTypeCnn3DLayers

public static InputPreProcessor getPreProcessorForInputTypeCnn3DLayers(InputType inputType,
                                                                       String layerName)

Utility method for determining the appropriate preprocessor for CNN layers, such as ConvolutionLayer and SubsamplingLayer

Parameters:

inputType - Input type to get the preprocessor for

Returns:

Null if no preprocessor is required; otherwise the appropriate preprocessor for the given input type

getPreProcessorForInputTypeCnnLayers

public static InputPreProcessor getPreProcessorForInputTypeCnnLayers(InputType inputType,
                                                                     String layerName)

Utility method for determining the appropriate preprocessor for CNN layers, such as ConvolutionLayer and SubsamplingLayer

Parameters:

inputType - Input type to get the preprocessor for

Returns:

Null if no preprocessor is required; otherwise the appropriate preprocessor for the given input type

getPreprocessorForInputTypeRnnLayers

public static InputPreProcessor getPreprocessorForInputTypeRnnLayers(InputType inputType,
                                                                     RNNFormat rnnDataFormat,
                                                                     String layerName)

convertMultipleTypes

public static void convertMultipleTypes(InputType[] vertexInputs)

Convert multiple types when multiple are found. Only handles simple obvious cases, otherwise errs on throwing an exception. Useful for multiple input vertices such as MergeVertex and ElementWiseVertex

Parameters:

vertexInputs - the input types to convert