EasyCV  0.9.36
Easy! Computer Vision
Variables
bootstrapping Namespace Reference

Variables

tuple trainset1 = easy.createRunSet( "corporate_logos" )
 
tuple trainer = easy.getTrainer( "BOW_Trainer")
 
tuple model1 = easy.train( trainer, trainset1 )
 
tuple testset1 = easy.createRunSet( "testImg", "UNPURPOSED" )
 
tuple detector = easy.getDetector( "BOW_Detector" )
 
tuple result1 = easy.detect( detector, model1, testset1 )
 
string reject_folder = easy.CVAC_DataDir+"/corporate_logos_round2/reject"
 
list nologos = ["TestKrFlag.jpg", "italia.jpg", "korean-american-flag.jpg", "TestUsFlag.jpg"]
 
tuple fname = os.path.join(root, filename)
 
string newf = reject_folder+"/"
 
list mapwithreject = trainset1['classmap']
 
tuple trainset2 = easy.createRunSet( "corporate_logos_round2", classmap=mapwithreject )
 
tuple model2 = easy.train( trainer, trainset2 )
 

Detailed Description

Easy!  mini tutorial
Repeatedly train and evaluate for efficient label use; bootstrap.
matz 6/21/2013

Variable Documentation

tuple bootstrapping.detector = easy.getDetector( "BOW_Detector" )
tuple bootstrapping.fname = os.path.join(root, filename)
list bootstrapping.mapwithreject = trainset1['classmap']
tuple bootstrapping.model1 = easy.train( trainer, trainset1 )
tuple bootstrapping.model2 = easy.train( trainer, trainset2 )
string bootstrapping.newf = reject_folder+"/"
list bootstrapping.nologos = ["TestKrFlag.jpg", "italia.jpg", "korean-american-flag.jpg", "TestUsFlag.jpg"]
string bootstrapping.reject_folder = easy.CVAC_DataDir+"/corporate_logos_round2/reject"
tuple bootstrapping.result1 = easy.detect( detector, model1, testset1 )
tuple bootstrapping.testset1 = easy.createRunSet( "testImg", "UNPURPOSED" )
tuple bootstrapping.trainer = easy.getTrainer( "BOW_Trainer")
tuple bootstrapping.trainset1 = easy.createRunSet( "corporate_logos" )
tuple bootstrapping.trainset2 = easy.createRunSet( "corporate_logos_round2", classmap=mapwithreject )