def gen(df):
"""
Wrapper around generator. Keras fit_generator requires looping generator.
:param df: dataflow instance
"""
while True:
for i in df.get_data():
yield i
if __name__ == '__main__':
# get the model
model = get_training_model(weight_decay)
# restore weights
last_epoch = restore_weights(weights_best_file, model)
# prepare generators
curr_dir = os.path.dirname(__file__)
annot_path_train = os.path.join(curr_dir, '../dataset/annotations/person_keypoints_train2017.json')
img_dir_train = os.path.abspath(os.path.join(curr_dir, '../dataset/train2017/'))
annot_path_val = os.path.join(curr_dir, '../dataset/annotations/person_keypoints_val2017.json')
img_dir_val = os.path.abspath(os.path.join(curr_dir, '../dataset/val2017/'))
# get dataflow of samples from training set and validation set (we use validation set for training as well)
:param df: dataflow instance
"""
while True:
for i in df.get_data():
yield i
if __name__ == '__main__':
parser = argparse.ArgumentParser(
description='Training codes for Keras Pose Estimation')
parser.add_argument('--model', default='cmu', help='model name')
args = parser.parse_args()
# get the model
if args.model == 'cmu':
model = cmu_model.get_training_model(weight_decay)
elif args.model == 'squeeze':
model = squeeze_model.get_training_model(weight_decay)
elif args.model == 'mobile':
model = mobile_model.get_training_model(weight_decay)
elif args.model == 'cmu-thin':
model = cmu_thin_model.get_training_model(weight_decay)
# restore weights
last_epoch = restore_weights(weights_best_file, model)
#last_epoch = restore_weights("../model/squeeze_imagenet.h5", model)
print(model.summary())
# prepare generators
curr_dir = os.path.dirname(__file__)