.TH "MPSNNDefaultPadding" 3 "Mon Jul 9 2018" "Version MetalPerformanceShaders-119.3" "MetalPerformanceShaders.framework" \" -*- nroff -*-
.ad l
.nh
.SH NAME
MPSNNDefaultPadding
.SH SYNOPSIS
.br
.PP
.PP
\fC#import <MPSNeuralNetworkTypes\&.h>\fP
.PP
Inherits NSObject, and \fB<MPSNNPadding>\fP\&.
.SS "Instance Methods"

.in +1c
.ti -1c
.RI "(NSString *__nonnull) \- \fBlabel\fP"
.br
.in -1c
.SS "Class Methods"

.in +1c
.ti -1c
.RI "(instancetype __nonnull) + \fBpaddingWithMethod:\fP"
.br
.ti -1c
.RI "(instancetype __nonnull) + \fBpaddingForTensorflowAveragePooling\fP"
.br
.ti -1c
.RI "(instancetype __nonnull) + \fBpaddingForTensorflowAveragePoolingValidOnly\fP"
.br
.in -1c
.SH "Method Documentation"
.PP 
.SS "\- (NSString * __nonnull) label "
Human readable description of what the padding policy does 
.SS "+ (instancetype __nonnull) paddingForTensorflowAveragePooling "
\fBA\fP padding policy that attempts to reproduce TensorFlow behavior for average pooling  Most TensorFlow padding is covered by the standard MPSNNPaddingMethod encodings\&. You can use +paddingWithMethod to get quick access to \fBMPSNNPadding\fP objects, when default filter behavior isn't enough\&. (It often is\&.) However, the edging for max pooling in TensorFlow is a bit unusual\&.
.PP
This padding method attempts to reproduce TensorFlow padding for average pooling\&. In addition to setting MPSNNPaddingMethodSizeSame | MPSNNPaddingMethodAlignCentered | MPSNNPaddingMethodAddRemainderToBottomRight, it also configures the filter to run with MPSImageEdgeModeClamp, which (as a special case for average pooling only), normalizes the sum of contributing samples to the area of valid contributing pixels only\&.
.PP
.PP
.nf
// Sample implementation for the tensorflowPoolingPaddingPolicy returned
 -(MPSNNPaddingMethod) paddingMethod{ return MPSNNPaddingMethodCustom | MPSNNPaddingMethodSizeSame; }

 -(MPSImageDescriptor * __nonnull) destinationImageDescriptorForSourceImages: (NSArray <MPSImage *> *__nonnull) sourceImages
                                                                sourceStates: (NSArray <MPSState *> * __nullable) sourceStates
                                                                   forKernel: (MPSKernel * __nonnull) kernel
                                                         suggestedDescriptor: (MPSImageDescriptor * __nonnull) inDescriptor
 {

    ((MPSCNNKernel *)kernel)\&.edgeMode = MPSImageEdgeModeClamp;

    return inDescriptor;
 }
.fi
.PP
 
.SS "+ (instancetype __nonnull) paddingForTensorflowAveragePoolingValidOnly "
Typical pooling padding policy for valid only mode 
.SS "+ (instancetype __nonnull) paddingWithMethod: (\fBMPSNNPaddingMethod\fP) method"
Fetch a well known object that implements a non-custom padding method  For custom padding methods, you will need to implement an object that conforms to the full \fBMPSNNPadding\fP protocol, including \fBNSSecureCoding\fP\&. 
.PP
\fBParameters:\fP
.RS 4
\fImethod\fP \fBA\fP MPSNNPaddingMethod 
.RE
.PP
\fBReturns:\fP
.RS 4
An object that implements <MPSNNPadding> for use with MPSNNGraphNodes\&. 
.RE
.PP


.SH "Author"
.PP 
Generated automatically by Doxygen for MetalPerformanceShaders\&.framework from the source code\&.