EasyCV  0.9.36
Easy! Computer Vision
Variables
bowDemo_ROC Namespace Reference

Variables

tuple stime = time.clock()
 
tuple trainsetPos = easy.createRunSet( 'corporate_logos' )
 
tuple trainsetNeg = easy.createRunSet( 'trainImg' )
 
tuple runset = cvac.RunSet()
 
string strTrainer = "BOW_Trainer"
 
string strDetector = "BOW_Detector"
 
list list_nWord = [5,10,15,20]
 
 doWithNegativeSample = True
 
list contenders = []
 With background data. More...
 
tuple c1 = evaluate.Contender("bowROC_binary_"+str(nWord))
 
tuple trainer = easy.getTrainer(c1.trainerString)
 
tuple trainerProps = easy.getTrainerProperties(trainer)
 
list rocData_optimal = []
 
tuple detectorData
 
tuple rocZip = easy.makeROCdata(rocData_optimal)
 
tuple detector = easy.getDetector( strDetector )
 
tuple detectorProps = easy.getDetectorProperties(detector)
 
string priority = "recall"
 
tuple results = easy.detect( detector, rocZip, runset, detectorProperties = detectorProps)
 
tuple opPoints = easy.getSensitivityOptions(rocZip)
 
tuple etime = time.clock()
 

Detailed Description

Generate ROC curve with Bag-of-Words algorithm.
Among optimal operating points, 
the best point and the corresponding detectorData will be returned  
generated by matz 7/16/2013
updated by k.lee May/2014

Variable Documentation

tuple bowDemo_ROC.c1 = evaluate.Contender("bowROC_binary_"+str(nWord))
list bowDemo_ROC.contenders = []

With background data.

Without background data.

tuple bowDemo_ROC.detector = easy.getDetector( strDetector )
tuple bowDemo_ROC.detectorData
Initial value:
1 = easy.train( trainer, runset, \
2  trainerProperties=contenders[idx].trainerProps )
tuple bowDemo_ROC.detectorProps = easy.getDetectorProperties(detector)
bowDemo_ROC.doWithNegativeSample = True
tuple bowDemo_ROC.etime = time.clock()
list bowDemo_ROC.list_nWord = [5,10,15,20]
tuple bowDemo_ROC.opPoints = easy.getSensitivityOptions(rocZip)
string bowDemo_ROC.priority = "recall"
tuple bowDemo_ROC.results = easy.detect( detector, rocZip, runset, detectorProperties = detectorProps)
list bowDemo_ROC.rocData_optimal = []
tuple bowDemo_ROC.rocZip = easy.makeROCdata(rocData_optimal)
tuple bowDemo_ROC.runset = cvac.RunSet()
tuple bowDemo_ROC.stime = time.clock()
string bowDemo_ROC.strDetector = "BOW_Detector"
string bowDemo_ROC.strTrainer = "BOW_Trainer"
list bowDemo_ROC.trainer = easy.getTrainer(c1.trainerString)
tuple bowDemo_ROC.trainerProps = easy.getTrainerProperties(trainer)
tuple bowDemo_ROC.trainsetNeg = easy.createRunSet( 'trainImg' )
tuple bowDemo_ROC.trainsetPos = easy.createRunSet( 'corporate_logos' )