Specifies the filler parameters used to create each Filler. More...

Inheritance diagram for MyCaffe.param.FillerParameter:

Public Types
enum	FillerType { CONSTANT , SEQUENCE , UNIFORM , GAUSSIAN , XAVIER , MSRA , POSITIVEUNITBALL , BILINEAR }
	Defines the type of filler. More...

enum	VarianceNorm { FAN_IN = 0 , FAN_OUT = 1 , AVERAGE = 2 }
	Defines the variance normalization. More...

Public Member Functions
	FillerParameter (string strType="gaussian", double dfVal=0.0, double dfMean=0.0, double dfStd=1.0)
	Filler parameter constructor More...

FillerParameter	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.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 string	GetFillerName (FillerType type)
	Queries the filler text name corresponding to the FillerType. More...

static FillerParameter	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
string	type `[getset]`
	Specifies the type of filler to use. More...

double	value `[getset]`
	Specifies the value used by 'constant' filler. More...

double	min `[getset]`
	Specifies the minimum value to use with the 'uniform' filler. More...

double	max `[getset]`
	Specifies the maximum value to use with the 'uniform' filler. More...

double	mean `[getset]`
	Specifies the mean value to use with the 'gaussian' filler. More...

double	std `[getset]`
	Specifies the standard deviation value to use with the 'gaussian' filler. More...

int	sparse `[getset]`
	Specifies the sparcity value to use with the 'guassian' filler. More...

VarianceNorm	variance_norm `[getset]`
	Specifies the variance normalization method to use with the 'xavier' and 'mrsa' fillers. More...

Detailed Description

Specifies the filler parameters used to create each Filler.

Definition at line 15 of file FillerParameter.cs.

Member Enumeration Documentation

◆ FillerType

enum MyCaffe.param.FillerParameter.FillerType

Defines the type of filler.

Enumerator
CONSTANT	The constant filler fills a blob with constant values.
SEQUENCE	The sequence filler fills a blob with values starting at the constant value and progressing in steps of 0.01. This filler is primarily used for testing.
UNIFORM	The uniform filler fills a blob with values from a uniform distribution. See also Uniform Distribution Wikipedia.
GAUSSIAN	The gaussian filler fills a blob with values from a gaussian distribution. See also Guassian Distribution Wikipedia.
XAVIER	The xavier filler fills a blob with values from a xavier distribution. See also Understanding the difficulty of training deep feedforward neural networks by Glorot, Xavier and Bengio, Yoshua, 2010.
MSRA	The msra filler fills a blob with values from a msra distribution. See also Learning hierarchical categories in deep neural networks by Saxe, Andrew M. and McClelland, James L. and Ganguli, Surya, 2013. Delving Deep into Rectifiers: Surpassing Human-Level Performance on ImageNet Classification by He, Kaiming and Zhang, Xiangyu and Ren, Shaoqing and Sun, Jian, 2015.
POSITIVEUNITBALL	The positive unit ball filler fills a blob with values from a positive unit ball distribution.
BILINEAR	The bilinear filler fills a blob with values from a bilinear distribution.

Definition at line 34 of file FillerParameter.cs.

◆ VarianceNorm

enum MyCaffe.param.FillerParameter.VarianceNorm

Defines the variance normalization.

Enumerator
FAN_IN	Specifies a fan-in variance normalization.
FAN_OUT	Specifies a fan-out variance normalization.
AVERAGE	Specifies an average variance normalization.

Definition at line 87 of file FillerParameter.cs.

Constructor & Destructor Documentation

◆ FillerParameter()

MyCaffe.param.FillerParameter.FillerParameter	(	string	strType = `"gaussian"`,
		double	dfVal = `0.0`,
		double	dfMean = `0.0`,
		double	dfStd = `1.0`
	)

Filler parameter constructor

NOTE: Caffe defaults to 'constant', however this causes models that do not specifically specify a filler to run with constant 0 filled weights. Using a 'gaussian' as the default fixes this and fills the weights with random numbers.

Parameters

strType	Optionally, specifies the type of filler to use. Default = "gaussian"
dfVal	Optionally, specifies the value. Default = 0.0
dfMean	Optionally, specifies the mean. Default = 0.0
dfStd	Optionally, specifies the standard deviation. Default = 1.0

Definition at line 116 of file FillerParameter.cs.

Member Function Documentation

◆ Clone()

FillerParameter MyCaffe.param.FillerParameter.Clone ( )

Creates a new copy of this instance of the parameter.

Returns: A new instance of this parameter is returned.

Definition at line 293 of file FillerParameter.cs.

◆ FromProto()

static FillerParameter MyCaffe.param.FillerParameter.FromProto ( RawProto rp )

static

Parses the parameter from a RawProto.

Parameters

rp	Specifies the RawProto to parse.

Returns: A new instance of the parameter is returned.

Definition at line 350 of file FillerParameter.cs.

◆ GetFillerName()

static string MyCaffe.param.FillerParameter.GetFillerName ( FillerType type )

static

Queries the filler text name corresponding to the FillerType.

Parameters

type	Specifies the FillerType.

Returns: The string associated with the FillerType is returned.

Definition at line 186 of file FillerParameter.cs.

◆ ToProto()

override RawProto MyCaffe.param.FillerParameter.ToProto ( string strName )

virtual

Convert the parameter into a RawProto.

Parameters

strName Specifies the name to associate with the RawProto.

Returns: The new RawProto is returned.

Implements MyCaffe.basecode.BaseParameter.

Definition at line 314 of file FillerParameter.cs.

Property Documentation

◆ max

double MyCaffe.param.FillerParameter.max

getset

Specifies the maximum value to use with the 'uniform' filler.

Definition at line 243 of file FillerParameter.cs.

◆ mean

double MyCaffe.param.FillerParameter.mean

getset

Specifies the mean value to use with the 'gaussian' filler.

Definition at line 253 of file FillerParameter.cs.

◆ min

double MyCaffe.param.FillerParameter.min

getset

Specifies the minimum value to use with the 'uniform' filler.

Definition at line 233 of file FillerParameter.cs.

◆ sparse

int MyCaffe.param.FillerParameter.sparse

getset

Specifies the sparcity value to use with the 'guassian' filler.

Definition at line 273 of file FillerParameter.cs.

◆ std

double MyCaffe.param.FillerParameter.std

getset

Specifies the standard deviation value to use with the 'gaussian' filler.

Definition at line 263 of file FillerParameter.cs.

◆ type

string MyCaffe.param.FillerParameter.type

getset

Specifies the type of filler to use.

Definition at line 131 of file FillerParameter.cs.

◆ value

double MyCaffe.param.FillerParameter.value

getset

Specifies the value used by 'constant' filler.

Definition at line 223 of file FillerParameter.cs.

◆ variance_norm

VarianceNorm MyCaffe.param.FillerParameter.variance_norm

getset

Specifies the variance normalization method to use with the 'xavier' and 'mrsa' fillers.

Definition at line 283 of file FillerParameter.cs.

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

C:/Data/Data/SS_Projects/Intelligence/GitHub/MyCaffe/MyCaffe/param/FillerParameter.cs

Public Types

Public Member Functions

Static Public Member Functions

Properties

Detailed Description

Member Enumeration Documentation

◆ FillerType

◆ VarianceNorm

Constructor & Destructor Documentation

◆ FillerParameter()

Member Function Documentation

◆ Clone()

◆ FromProto()

◆ GetFillerName()

◆ ToProto()

Property Documentation

◆ max

◆ mean

◆ min

◆ sparse

◆ std

◆ type

◆ value

◆ variance_norm