本文整理匯總了Python中slim.datasets.dataset_utils.image_to_tfexample方法的典型用法代碼示例。如果您正苦於以下問題：Python dataset_utils.image_to_tfexample方法的具體用法？Python dataset_utils.image_to_tfexample怎麽用？Python dataset_utils.image_to_tfexample使用的例子？那麽, 這裏精選的方法代碼示例或許可以為您提供幫助。您也可以進一步了解該方法所在類slim.datasets.dataset_utils的用法示例。
在下文中一共展示了dataset_utils.image_to_tfexample方法的1個代碼示例，這些例子默認根據受歡迎程度排序。您可以為喜歡或者感覺有用的代碼點讚，您的評價將有助於係統推薦出更棒的Python代碼示例。
示例1: _convert_dataset
# 需要導入模塊: from slim.datasets import dataset_utils [as 別名]
# 或者: from slim.datasets.dataset_utils import image_to_tfexample [as 別名]
def _convert_dataset(split_name, filenames, filename_to_class_id, dataset_dir):
"""Converts the given filenames to a TFRecord dataset.
Args:
split_name: The name of the dataset, either 'train' or 'valid'.
filenames: A list of absolute paths to png images.
filename_to_class_id: A dictionary from filenames (strings) to class ids
(integers).
dataset_dir: The directory where the converted datasets are stored.
"""
print('Converting the {} split.'.format(split_name))
# Train and validation splits are both in the train directory.
if split_name in ['train', 'valid']:
png_directory = os.path.join(dataset_dir, 'mnist_m', 'mnist_m_train')
elif split_name == 'test':
png_directory = os.path.join(dataset_dir, 'mnist_m', 'mnist_m_test')
with tf.Graph().as_default():
image_reader = ImageReader()
with tf.Session('') as sess:
output_filename = _get_output_filename(dataset_dir, split_name)
with tf.python_io.TFRecordWriter(output_filename) as tfrecord_writer:
for filename in filenames:
# Read the filename:
image_data = tf.gfile.FastGFile(
os.path.join(png_directory, filename), 'r').read()
height, width = image_reader.read_image_dims(sess, image_data)
class_id = filename_to_class_id[filename]
example = dataset_utils.image_to_tfexample(image_data, 'png', height,
width, class_id)
tfrecord_writer.write(example.SerializeToString())
sys.stdout.write('\n')
sys.stdout.flush()