app-combineOptimalGains.C File Reference

#include "Channels/OptimalGains.H"
#include "Component/ModelManager.H"
#include "Component/ParamMap.H"
#include "Component/ModelOptionDef.H"
#include "Neuro/NeuroOpts.H"
#include "Util/sformat.H"

Include dependency graph for app-combineOptimalGains.C:

Go to the source code of this file.

Functions
rutz::shared_ptr< ParamMap >	combineParamMaps (std::vector< rutz::shared_ptr< ParamMap > > pmaps, const uint indent)
int	main (const int argc, const char **argv)
	Combine several stsd.pmap files into a gains.pmap file.
Variables
static const ModelOptionDef	OPT_SaveTo

---

Detailed Description

combine optimal gains files

Definition in file app-combineOptimalGains.C.

---

Function Documentation

int main	(	const int	argc,
		const char **	argv
	)

Combine several stsd.pmap files into a gains.pmap file.

The stsd.pmap files should be obtained by running something like:

ezvision --in=xmlfile:testfile.xml --out=display --pfc-type=OG --vc-type=Std --vc-chans=GNO --stsd-filename=stsd.pmap --nouse-older-version

which will compute the salience of target and distractor and save those to a ParamMap. Once you have several of these (or just one), you can use the present program to generate an optimal gains ParamMap (it is written to stdout). Finally, you can use these gains for biased saliency computations using a PrefrontalCortexGS, e.g.:

ezvision --in=image.png --out=display -X --pfc-type=GS --vc-type=Std --vc-chans=GNO --gains-filename=gains.pmap --nouse-older-version

Definition at line 138 of file app-combineOptimalGains.C.

References ModelManager::getExtraArg(), OModelParam< T >::getVal(), ParamMap::loadPmapFile(), MYLOGVERB, ModelManager::numExtraArgs(), and ModelManager::parseCommandLine().

---

Variable Documentation

const ModelOptionDef OPT_SaveTo [static]

Initial value:

  { MODOPT_ARG(std::string), "SaveTo", &MOC_DISPLAY, OPTEXP_SAVE,
    "Save output gains to file if name specified (rather than stdout)",
    "save-to", '\0', "", "" }

Definition at line 45 of file app-combineOptimalGains.C.