detector = Detector(learning_component)
for patch in positive_patches[1:max_train_count]:
learning_component.update_positives(patch)
for patch in negative_patches[:5*max_train_count]:
learning_component.update_negatives(patch)
print "Update training set:", time.time() - start_time
print
start_time = time.time()
detector.ensemble_classifier.relearn()
print "Learn detector:", time.time() - start_time
print
for param in xrange(10):
print "Param:", param/10.0
detector.threshold_patch_variance = 0.3
detector.threshold_ensemble = param/10.0
detector.nearest_neighbor_classifier.tetta = 0.4
start_time = time.time()
tp = 0
tn = 0
fp = 0
fn = 0
for patch in positive_patches[-max_test_count:]:
if detector.predict_patch(patch) > 0.5:
tp += 1
else:
fn += 1
for patch in negative_patches[-max_test_count:]:
if detector.predict_patch(patch) > 0.5:
fp += 1
