Main Page | Modules | Namespace List | Class Hierarchy | Alphabetical List | Class List | Directories | File List | Namespace Members | Class Members | File Members | Related Pages

WinnerTakeAll Class Reference

#include <Neuro/WinnerTakeAll.H>

Inheritance diagram for WinnerTakeAll:

Inheritance graph
[legend]
Collaboration diagram for WinnerTakeAll:

Collaboration graph
[legend]
List of all members.

Detailed Description

The winner-take-all base class.

This is a 2D winner-take-all, a.k.a., spatial maximum detector. It is just a base class with virtual function definitions. Various winner-take-all variants are available below, which provide implementations for these virtual functions. The WTA receives input through the input() function, and its current state (membrane potential of neurons) can be retrieved using getV(). Various inplementations will typically derive from class Image and inherit all the standard methods of class Image.

See also WinnerTakeAllAdapter for an intermediate base class that provides the getV() output functions. Why are those functions not in this base class? The reason is that Brain only requires the input interface of WinnerTakeAll, and separating the output interface allows us to make a stub class that has just no-op input functions.

Definition at line 77 of file WinnerTakeAll.H.

Public Member Functions

Constructor, destructor, and reset
 WinnerTakeAll (OptionManager &mgr, const std::string &descrName, const std::string &tagName)
 Ininitialized constructor.
virtual ~WinnerTakeAll ()
 Destructor.


Constructor & Destructor Documentation

WinnerTakeAll::WinnerTakeAll OptionManager mgr,
const std::string descrName,
const std::string tagName
 

Ininitialized constructor.

The SM will be resized and initialized the first time input() is called

Definition at line 64 of file WinnerTakeAll.C.

References GVX_TRACE.

WinnerTakeAll::~WinnerTakeAll  )  [virtual]
 

Destructor.

Definition at line 73 of file WinnerTakeAll.C.

References GVX_TRACE.


The documentation for this class was generated from the following files:
Generated on Sat Nov 7 13:36:25 2009 for iLab Neuromorphic Vision Toolkit by  doxygen 1.4.4