ONE - On-device Neural Engine
Loading...
Searching...
No Matches
loco::FilterShape Class Referencefinal

Filter Shape. More...

#include <FilterShape.h>

Public Member Functions

 FilterShape ()=default
 
const Dimensioncount (void) const
 
Dimensioncount (void)
 
const Dimensiondepth (void) const
 
Dimensiondepth (void)
 
const Dimensionheight (void) const
 
Dimensionheight (void)
 
const Dimensionwidth (void) const
 
Dimensionwidth (void)
 

Detailed Description

Filter Shape.

This class describes the shape of filter, which is an input of 2D convolutional operations (e.g. Convolution).

count() refers to the number of 3D weight in a filter depth() refers to the depth of 3D weights height() refers to the height of 3D weights width() refers to the width of 3D weights

NOTE

The definition of FilterShape is almost same as that of FeatureShape, but loco distinguishes FeatureShape and FilterShape in class-level in order to prevent potential errors by type check.

Definition at line 42 of file FilterShape.h.

Constructor & Destructor Documentation

◆ FilterShape()

loco::FilterShape::FilterShape ( )
default

Member Function Documentation

◆ count() [1/2]

Dimension & loco::FilterShape::count ( void  )
inline

Definition at line 49 of file FilterShape.h.

49{ return _count; }

◆ count() [2/2]

◆ depth() [1/2]

Dimension & loco::FilterShape::depth ( void  )
inline

Definition at line 52 of file FilterShape.h.

52{ return _depth; }

◆ depth() [2/2]

◆ height() [1/2]

Dimension & loco::FilterShape::height ( void  )
inline

Definition at line 55 of file FilterShape.h.

55{ return _height; }

◆ height() [2/2]

◆ width() [1/2]

Dimension & loco::FilterShape::width ( void  )
inline

Definition at line 58 of file FilterShape.h.

58{ return _width; }

◆ width() [2/2]


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