def rebalanced_distorted_inputs(pos_sample_dir):
"""Construct distorted input with augmented positive samples for specific
classes (symmetry).
Args:
pos_sample_dir: string. Directory of positive data.
Returns:
"""
if not FLAGS.data_dir:
raise ValueError('Please supply a data_dir')
data_dir = os.path.join(FLAGS.data_dir, 'fbb_output')
coefs1, images1, labels1 = nn_input.distorted_inputs(data_dir=data_dir,
batch_size=int(FLAGS.batch_size / 2))
coefs2, images2, labels2 = nn_input.distorted_inputs(data_dir=pos_sample_dir,
batch_size=int(FLAGS.batch_size / 2),
input_type='aug')
return tf.concat(0, [coefs1, coefs2]), \
tf.concat(0, [images1, images2]), \
tf.concat(0, [labels1, labels2])
def distorted_inputs(shuffle=True, num_threads=16, nodistort=False):
"""Construct distorted input for CIFAR training using the Reader ops.
Returns:
images: Images. 4D tensor of [batch_size, IMAGE_SIZE, IMAGE_SIZE, 3] size.
labels: Labels. 1D tensor of [batch_size] size.
Raises:
ValueError: If no data_dir
"""
# if not FLAGS.data_dir:
# raise ValueError('Please supply a data_dir')
# data_dir = os.path.join(FLAGS.data_dir, 'fbb_output')
return nn_input.distorted_inputs(run_config=run_config,
batch_size=FLAGS.batch_size,
shuffle=shuffle,
num_threads=num_threads,
nodistort=nodistort)
