MyCaffe  1.12.2.41
Deep learning software for Windows C# programmers.
MyCaffe.common.ResultCollection Class Reference

The ResultCollection contains the result of a given CaffeControl::Run. More...

Public Types

enum  RESULT_TYPE {
  NONE , PROBABILITIES , DISTANCES , MULTIBOX ,
  TEMPORAL
}
 Defines the type of result. More...
 

Public Member Functions

 ResultCollection (List< Result > rgResults, LayerParameter.LayerType outputLayerType)
 The ResultCollection constructor. More...
 
List< double > GetEncoding ()
 Returns the data encoding values. More...
 
void SetLabels (List< LabelDescriptor > rgLabels)
 Sets the label names in the label dictionary lookup. More...
 
override string ToString ()
 Returns a string representation of the results. More...
 
Image ToImage (ColorMapper clrMap)
 Converts the result collection into an image. More...
 

Static Public Member Functions

static RESULT_TYPE GetResultType (LayerParameter.LayerType type)
 Get the result type based on the layer-type used. More...
 

Properties

RESULT_TYPE ResultType [get]
 Returns the result type of the result data: PROBABILITIES (Sigmoid), DISTANCES (Decode), or NONE (Unknown). More...
 
List< ResultResultsOriginal [get]
 Returns the original results. More...
 
List< ResultResultsSorted [get]
 Returns the original results in sorted order. More...
 
int DetectedLabelMaxSignal [get]
 Returns the detected label with the maximum signal. More...
 
double DetectedLabelOutputMaxSignal [get]
 Returns the detected label output with the maximum signal. More...
 
int DetectedLabelMinSignal [get]
 Returns the detected label with the minimum signal. More...
 
double DetectedLabelOutputMinSignal [get]
 Returns the detected label output of the label with the minimum signal. More...
 
int DetectedLabel [get]
 Returns the detected label depending on the result type (distance or probability) with a default type of probability (max label signal) used. More...
 
double DetectedLabelOutput [get]
 Returns the detected label output depending on the result type (distance or probability) with a default type of probability (max label signal) used. More...
 
Dictionary< int, string > Labels [get]
 Returns the dictionary lookup of the labels and their names. More...
 

Detailed Description

The ResultCollection contains the result of a given CaffeControl::Run.

Definition at line 16 of file ResultCollection.cs.

Member Enumeration Documentation

◆ RESULT_TYPE

Defines the type of result.

Enumerator
NONE 

Specifies that no result type for the data.

PROBABILITIES 

Specifies that the results represent probabilities.

DISTANCES 

Specifies that the results represent distances.

MULTIBOX 

Specifies that the results represent multibox results.

TEMPORAL 

Specifies that the results represent temporal results.

Definition at line 26 of file ResultCollection.cs.

Constructor & Destructor Documentation

◆ ResultCollection()

MyCaffe.common.ResultCollection.ResultCollection ( List< Result rgResults,
LayerParameter.LayerType  outputLayerType 
)

The ResultCollection constructor.

Parameters
rgResultsSpecifies the results.
outputLayerTypeSpecifies the output layer type.

Definition at line 55 of file ResultCollection.cs.

Member Function Documentation

◆ GetEncoding()

List< double > MyCaffe.common.ResultCollection.GetEncoding ( )

Returns the data encoding values.

Definition at line 102 of file ResultCollection.cs.

◆ GetResultType()

static RESULT_TYPE MyCaffe.common.ResultCollection.GetResultType ( LayerParameter.LayerType  type)
static

Get the result type based on the layer-type used.

Parameters
typeSpecifies the layer-type used.
Returns
The result type of: PROBABILITIES, DECODE or MULTIBOX is returned.

Definition at line 73 of file ResultCollection.cs.

◆ SetLabels()

void MyCaffe.common.ResultCollection.SetLabels ( List< LabelDescriptor rgLabels)

Sets the label names in the label dictionary lookup.

Parameters
rgLabels

Definition at line 222 of file ResultCollection.cs.

◆ ToImage()

Image MyCaffe.common.ResultCollection.ToImage ( ColorMapper  clrMap)

Converts the result collection into an image.

Parameters
clrMapOptionally, specifies a colormap to use.
Returns
The image respresentation of the result collection is returned.

Definition at line 276 of file ResultCollection.cs.

◆ ToString()

override string MyCaffe.common.ResultCollection.ToString ( )

Returns a string representation of the results.

Returns

Definition at line 237 of file ResultCollection.cs.

Property Documentation

◆ DetectedLabel

int MyCaffe.common.ResultCollection.DetectedLabel
get

Returns the detected label depending on the result type (distance or probability) with a default type of probability (max label signal) used.

Definition at line 185 of file ResultCollection.cs.

◆ DetectedLabelMaxSignal

int MyCaffe.common.ResultCollection.DetectedLabelMaxSignal
get

Returns the detected label with the maximum signal.

The maximum signal label is used to detect the output from a SoftMax where each label is given a probability and the label with the highest probability is the detected label.

Definition at line 138 of file ResultCollection.cs.

◆ DetectedLabelMinSignal

int MyCaffe.common.ResultCollection.DetectedLabelMinSignal
get

Returns the detected label with the minimum signal.

The minimum signal label is used to detect the output from a Decode alyer where each label is given the distance from which the data's encoding is from the centroid of the label - the encoding with the minimum distance signifies the detected label.

Definition at line 164 of file ResultCollection.cs.

◆ DetectedLabelOutput

double MyCaffe.common.ResultCollection.DetectedLabelOutput
get

Returns the detected label output depending on the result type (distance or probability) with a default type of probability (max label signal) used.

Definition at line 199 of file ResultCollection.cs.

◆ DetectedLabelOutputMaxSignal

double MyCaffe.common.ResultCollection.DetectedLabelOutputMaxSignal
get

Returns the detected label output with the maximum signal.

The maximum signal label is used to detect the output from a SoftMax where each label is given a probability and the label with the highest probability is the detected label.

Definition at line 151 of file ResultCollection.cs.

◆ DetectedLabelOutputMinSignal

double MyCaffe.common.ResultCollection.DetectedLabelOutputMinSignal
get

Returns the detected label output of the label with the minimum signal.

The minimum signal label is used to detect the output from a Decode alyer where each label is given the distance from which the data's encoding is from the centroid of the label - the encoding with the minimum distance signifies the detected label.

Definition at line 177 of file ResultCollection.cs.

◆ Labels

Dictionary<int, string> MyCaffe.common.ResultCollection.Labels
get

Returns the dictionary lookup of the labels and their names.

Definition at line 213 of file ResultCollection.cs.

◆ ResultsOriginal

List<Result> MyCaffe.common.ResultCollection.ResultsOriginal
get

Returns the original results.

Definition at line 117 of file ResultCollection.cs.

◆ ResultsSorted

List<Result> MyCaffe.common.ResultCollection.ResultsSorted
get

Returns the original results in sorted order.

Definition at line 125 of file ResultCollection.cs.

◆ ResultType

RESULT_TYPE MyCaffe.common.ResultCollection.ResultType
get

Returns the result type of the result data: PROBABILITIES (Sigmoid), DISTANCES (Decode), or NONE (Unknown).

Definition at line 94 of file ResultCollection.cs.


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