MyCaffe  1.12.2.41
Deep learning software for Windows C# programmers.
MyCaffe.param.BatchNormParameter Class Reference

Specifies the parameters for the BatchNormLayer. More...

Inheritance diagram for MyCaffe.param.BatchNormParameter:
MyCaffe.param.EngineParameter MyCaffe.param.LayerParameterBase MyCaffe.basecode.BaseParameter MyCaffe.basecode.IBinaryPersist

Public Member Functions

 BatchNormParameter ()
 Constructor for the parameter. More...
 
string useCaffeReason ()
 Returns the reason that Caffe version was used instead of NVIDIA's cuDnn. More...
 
bool useCudnn ()
 Queries whether or not to use NVIDIA's cuDnn. More...
 
override object Load (System.IO.BinaryReader br, bool bNewInstance=true)
 Load the parameter from a binary reader. More...
 
override void Copy (LayerParameterBase src)
 Copy on parameter to another. More...
 
override LayerParameterBase Clone ()
 Creates a new copy of this instance of the parameter. More...
 
override RawProto ToProto (string strName)
 Convert the parameter into a RawProto. More...
 
- Public Member Functions inherited from MyCaffe.param.EngineParameter
 EngineParameter ()
 Constructor for the parameter. More...
 
override object Load (System.IO.BinaryReader br, bool bNewInstance=true)
 Load the parameter from a binary reader. More...
 
override void Copy (LayerParameterBase src)
 Copy on parameter to another. More...
 
override LayerParameterBase Clone ()
 Creates a new copy of this instance of the parameter. More...
 
override RawProto ToProto (string strName)
 Convert the parameter into a RawProto. More...
 
- Public Member Functions inherited from MyCaffe.param.LayerParameterBase
 LayerParameterBase ()
 Constructor for the parameter. More...
 
virtual string PrepareRunModelInputs ()
 This method gives derivative classes a chance specify model inputs required by the run model. More...
 
virtual void PrepareRunModel (LayerParameter p)
 This method gives derivative classes a chance to prepare the layer for a run-model. More...
 
void Save (BinaryWriter bw)
 Save this parameter to a binary writer. More...
 
abstract object Load (BinaryReader br, bool bNewInstance=true)
 Load the parameter from a binary reader. More...
 
- Public Member Functions inherited from MyCaffe.basecode.BaseParameter
 BaseParameter ()
 Constructor for the parameter. More...
 
virtual bool Compare (BaseParameter p)
 Compare this parameter to another parameter. More...
 

Static Public Member Functions

static new BatchNormParameter FromProto (RawProto rp)
 Parses the parameter from a RawProto. More...
 
- Static Public Member Functions inherited from MyCaffe.param.EngineParameter
static EngineParameter FromProto (RawProto rp)
 Parses the parameter from a RawProto. More...
 
- Static Public Member Functions inherited from MyCaffe.basecode.BaseParameter
static double ParseDouble (string strVal)
 Parse double values using the US culture if the decimal separator = '.', then using the native culture, and if then lastly trying the US culture to handle prototypes containing '.' as the separator, yet parsed in a culture that does not use '.' as a decimal. More...
 
static bool TryParse (string strVal, out double df)
 Parse double values using the US culture if the decimal separator = '.', then using the native culture, and if then lastly trying the US culture to handle prototypes containing '.' as the separator, yet parsed in a culture that does not use '.' as a decimal. More...
 
static float ParseFloat (string strVal)
 Parse float values using the US culture if the decimal separator = '.', then using the native culture, and if then lastly trying the US culture to handle prototypes containing '.' as the separator, yet parsed in a culture that does not use '.' as a decimal. More...
 
static bool TryParse (string strVal, out float f)
 Parse doufloatble values using the US culture if the decimal separator = '.', then using the native culture, and if then lastly trying the US culture to handle prototypes containing '.' as the separator, yet parsed in a culture that does not use '.' as a decimal. More...
 

Properties

bool scale_bias [getset]
 Specifies to use the scale and bias terms, otherwise the scale = 1 and bias = 0 are used to form an identity operation. More...
 
FillerParameter scale_filler [getset]
 Specifies the scale filler used to fill the scale value. If null, a constant(1) filler is used. More...
 
FillerParameter bias_filler [getset]
 Specifies the bias filler used to file the bias value. If null, a constant(0) filler is used. More...
 
bool? use_global_stats [getset]
 If false, normalization is performed over the current mini-batch and global statistics are accumulated (but not yet used) by a moving average. If true, those accumulated mean and variance values are used for the normalization. By default, this is set to false when the network is in Phase.TRAINING, and true when the network is in the Phase.TESTING mode. More...
 
double moving_average_fraction [getset]
 Specifies how much the moving average decays each iteration. Smaller values make the moving average decay faster, giving more weight to the recent values. More...
 
double eps [getset]
 Specifies a small value to add to the variance estimate so that we don't divide by zero. More...
 
- Properties inherited from MyCaffe.param.EngineParameter
Engine engine [getset]
 Specifies the Engine in use. More...
 

Additional Inherited Members

- Public Types inherited from MyCaffe.param.EngineParameter
enum  Engine { DEFAULT = 0 , CAFFE = 1 , CUDNN = 2 }
 Defines the type of engine to use. More...
 
- Public Types inherited from MyCaffe.param.LayerParameterBase
enum  LABEL_TYPE { NONE , SINGLE , MULTIPLE , ONLY_ONE }
 Defines the label type. More...
 

Detailed Description

Specifies the parameters for the BatchNormLayer.

See also
Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift by Sergey Ioffe and Christian Szegedy, 2015.
In Defense of the Triplet Loss for Person Re-Identification by Alexander Hermans, Lucas Beyer, and Bastian Leibe, 2017.
Layer Normalization by Jimmy Lei Ba, and Jamie Ryan Kiros, and Geoffrey E. Hinton, 2016.

Definition at line 20 of file BatchNormParameter.cs.

Constructor & Destructor Documentation

◆ BatchNormParameter()

MyCaffe.param.BatchNormParameter.BatchNormParameter ( )

Constructor for the parameter.

Definition at line 30 of file BatchNormParameter.cs.

Member Function Documentation

◆ Clone()

override LayerParameterBase MyCaffe.param.BatchNormParameter.Clone ( )
virtual

Creates a new copy of this instance of the parameter.

Returns
A new instance of this parameter is returned.

Implements MyCaffe.param.LayerParameterBase.

Definition at line 180 of file BatchNormParameter.cs.

◆ Copy()

override void MyCaffe.param.BatchNormParameter.Copy ( LayerParameterBase  src)
virtual

Copy on parameter to another.

Parameters
srcSpecifies the parameter to copy.

Implements MyCaffe.param.LayerParameterBase.

Definition at line 154 of file BatchNormParameter.cs.

◆ FromProto()

static new BatchNormParameter MyCaffe.param.BatchNormParameter.FromProto ( RawProto  rp)
static

Parses the parameter from a RawProto.

Parameters
rpSpecifies the RawProto to parse.
Returns
A new instance of the parameter is returned.

Definition at line 227 of file BatchNormParameter.cs.

◆ Load()

override object MyCaffe.param.BatchNormParameter.Load ( System.IO.BinaryReader  br,
bool  bNewInstance = true 
)

Load the parameter from a binary reader.

Parameters
brSpecifies the binary reader.
bNewInstanceWhen true a new instance is created (the default), otherwise the existing instance is loaded from the binary reader.
Returns
Returns an instance of the parameter.

Definition at line 142 of file BatchNormParameter.cs.

◆ ToProto()

override RawProto MyCaffe.param.BatchNormParameter.ToProto ( string  strName)
virtual

Convert the parameter into a RawProto.

Parameters
strNameSpecifies the name to associate with the RawProto.
Returns
The new RawProto is returned.

Implements MyCaffe.basecode.BaseParameter.

Definition at line 192 of file BatchNormParameter.cs.

◆ useCaffeReason()

string MyCaffe.param.BatchNormParameter.useCaffeReason ( )

Returns the reason that Caffe version was used instead of NVIDIA's cuDnn.

Returns

Definition at line 38 of file BatchNormParameter.cs.

◆ useCudnn()

bool MyCaffe.param.BatchNormParameter.useCudnn ( )

Queries whether or not to use NVIDIA's cuDnn.

Returns
Returns true when cuDnn is to be used, false otherwise.

Definition at line 50 of file BatchNormParameter.cs.

Property Documentation

◆ bias_filler

FillerParameter MyCaffe.param.BatchNormParameter.bias_filler
getset

Specifies the bias filler used to file the bias value. If null, a constant(0) filler is used.

NOTE: Currently the scale_bias is only used by the CUDNN engine.

Definition at line 92 of file BatchNormParameter.cs.

◆ eps

double MyCaffe.param.BatchNormParameter.eps
getset

Specifies a small value to add to the variance estimate so that we don't divide by zero.

Definition at line 135 of file BatchNormParameter.cs.

◆ moving_average_fraction

double MyCaffe.param.BatchNormParameter.moving_average_fraction
getset

Specifies how much the moving average decays each iteration. Smaller values make the moving average decay faster, giving more weight to the recent values.

Each iteration updates the moving average $_{t-1}$ with the current mean $ Y_t $ by $ S_t = (1-\beta)Y_t + \beta \cdot S_{t-1} $, where $ \beta $ is the moving average fraction parameter.

Definition at line 124 of file BatchNormParameter.cs.

◆ scale_bias

bool MyCaffe.param.BatchNormParameter.scale_bias
getset

Specifies to use the scale and bias terms, otherwise the scale = 1 and bias = 0 are used to form an identity operation.

NOTE: Currently the scale_bias is only used by the CUDNN engine.

Definition at line 66 of file BatchNormParameter.cs.

◆ scale_filler

FillerParameter MyCaffe.param.BatchNormParameter.scale_filler
getset

Specifies the scale filler used to fill the scale value. If null, a constant(1) filler is used.

NOTE: Currently the scale_bias is only used by the CUDNN engine.

Definition at line 79 of file BatchNormParameter.cs.

◆ use_global_stats

bool? MyCaffe.param.BatchNormParameter.use_global_stats
getset

If false, normalization is performed over the current mini-batch and global statistics are accumulated (but not yet used) by a moving average. If true, those accumulated mean and variance values are used for the normalization. By default, this is set to false when the network is in Phase.TRAINING, and true when the network is in the Phase.TESTING mode.

Definition at line 108 of file BatchNormParameter.cs.


The documentation for this class was generated from the following file: