def main():
global gl_emotions
global gd_setup
global gd_config
gd_setup, gd_config = helper.load_setup()
gl_emotions = gd_config['emotions']
run_main_loop(gd_setup['testSetupPath'])
print "Task completed"
def run_all():
d_setup, __ = helper.load_setup()
if (len(d_setup)) > 0:
print "Creating Evaluation Setup: {0}".format(
d_setup['evaluationSetupPath'])
evaluation_setup_creator.main()
print "Training Classifier: {0}".format(d_setup['classifierPath'])
train.main()
print "Creating Test Setup: {0}".format(d_setup['testSetupPath'])
create_test_setup.main()
print "Computing Metrics for: {0}".format(d_setup['classifierPath'])
compute_metrics_calculation.main()
print 'Task Completed'
def main():
d_setup, __ = helper.load_setup()
ll_test_files = glob.glob(
path.join(d_setup['evaluationSetupPath'], '*test.csv'))
v_dest = d_setup['testSetupPath']
helper.create_directory(v_dest)
fold_count = 0
for fold in ll_test_files:
v_current_destination = path.join(v_dest, 'fold{0}'.format(fold_count))
helper.create_directory(v_current_destination)
copyfile(fold, path.join(v_current_destination,
'all_participants.csv'))
fold_count += 1
def main():
global emotions
gd_setup, __ = helper.load_setup()
emotions = gd_setup['emotionList']
v_basepath = gd_setup['evaluationSetupPath']
v_sourcepath = gd_setup['sourceFilesPath']
v_folds = gd_setup['evaluationFolds']
helper.create_directory((v_basepath))
create_participants_list(pv_basepath=v_basepath,
pv_sourcepath=v_sourcepath)
# create_participants_list(pv_basepath="test_setup2", pv_sourcepath="Data\\sorted_set_testing2")
for i in xrange(v_folds):
create_evaluation_setup(pv_basepath=v_basepath,
pv_sourcepath=v_sourcepath,
pv_fold_number=i,
limit_to_min=False)
# Update yaml dict
yaml_dict['basepath'] = v_basepath
yaml_dict['sourcepath'] = v_sourcepath
yaml_dict['folds'] = v_folds
yaml_dict['emotions'] = emotions
save_to_yaml(v_basepath)
import cv2
import predictor
import overlay_helper
import helper
import glob
from os import path
d_setup, d_config = helper.load_setup()
def get_image_list(pv_path):
ll_files = []
for img_format in ('*.png', '*.jpg', '*.jpeg'):
ll_files.extend(glob.glob(path.join(pv_path, img_format)))
return ll_files
def process(pv_path):
ll_image_files = get_image_list(pv_path)
for img_path in ll_image_files:
if '_result' in img_path:
continue
print img_path
# Read image
img = cv2.imread(img_path)
# Predict face emotions on frame
b_found, ll_pred, ll_pred_label, llol_pred_prob, landmarks = predictor.predict_image(
img)
def main():
global gd_setup
gd_setup, __ = helper.load_setup()
train(gd_setup['evaluationSetupPath'])
fm.json_save_file(
path.join(gd_setup['evaluationSetupPath'], 'results.json'), json_dict)