Python preprocessing.decode_image方法代码示例

# 需要导入模块: import preprocessing [as 别名]
# 或者: from preprocessing import decode_image [as 别名]
def parse_sequence_to_svtcn_batch(
    serialized_example, preprocess_fn, is_training, num_views, batch_size):
  """Parses a serialized sequence example into a batch of SVTCN data."""
  _, views, seq_len = parse_sequence_example(serialized_example, num_views)
  # Get svtcn indices.
  time_indices, view_indices = get_svtcn_indices(seq_len, batch_size, num_views)
  combined_indices = tf.concat(
      [tf.expand_dims(view_indices, 1),
       tf.expand_dims(time_indices, 1)], 1)

  # Gather the image strings.
  images = tf.gather_nd(views, combined_indices)

  # Decode images.
  images = tf.map_fn(preprocessing.decode_image, images, dtype=tf.float32)

  # Concatenate anchor and postitive images, preprocess the batch.
  preprocessed = preprocess_fn(images, is_training)

  return preprocessed, images, time_indices 

# 需要导入模块: import preprocessing [as 别名]
# 或者: from preprocessing import decode_image [as 别名]
def parse_labeled_example(
    example_proto, view_index, preprocess_fn, image_attr_keys, label_attr_keys):
  """Parses a labeled test example from a specified view.

  Args:
    example_proto: A scalar string Tensor.
    view_index: Int, index on which view to parse.
    preprocess_fn: A function with the signature (raw_images, is_training) ->
      preprocessed_images, where raw_images is a 4-D float32 image `Tensor`
      of raw images, is_training is a Boolean describing if we're in training,
      and preprocessed_images is a 4-D float32 image `Tensor` holding
      preprocessed images.
    image_attr_keys: List of Strings, names for image keys.
    label_attr_keys: List of Strings, names for label attributes.
  Returns:
    data: A tuple of images, attributes and tasks `Tensors`.
  """
  features = {}
  for attr_key in image_attr_keys:
    features[attr_key] = tf.FixedLenFeature((), tf.string)
  for attr_key in label_attr_keys:
    features[attr_key] = tf.FixedLenFeature((), tf.int64)
  parsed_features = tf.parse_single_example(example_proto, features)
  image_only_keys = [i for i in image_attr_keys if 'image' in i]
  view_image_key = image_only_keys[view_index]
  image = preprocessing.decode_image(parsed_features[view_image_key])
  preprocessed = preprocess_fn(image, is_training=False)
  attributes = [parsed_features[k] for k in label_attr_keys]
  task = parsed_features['task']
  return tuple([preprocessed] + attributes + [task]) 

# 需要导入模块: import preprocessing [as 别名]
# 或者: from preprocessing import decode_image [as 别名]
def parse_sequence_to_pairs_batch(
    serialized_example, preprocess_fn, is_training, num_views, batch_size,
    window):
  """Parses a serialized sequence example into a batch of preprocessed data.

  Args:
    serialized_example: A serialized SequenceExample.
    preprocess_fn: A function with the signature (raw_images, is_training) ->
      preprocessed_images.
    is_training: Boolean, whether or not we're in training.
    num_views: Int, the number of simultaneous viewpoints at each timestep in
      the dataset.
    batch_size: Int, size of the batch to get.
    window: Int, only take pairs from a maximium window of this size.
  Returns:
    preprocessed: A 4-D float32 `Tensor` holding preprocessed images.
    anchor_images: A 4-D float32 `Tensor` holding raw anchor images.
    pos_images: A 4-D float32 `Tensor` holding raw positive images.
  """
  _, views, seq_len = parse_sequence_example(serialized_example, num_views)

  # Get random (anchor, positive) timestep and viewpoint indices.
  num_pairs = batch_size // 2
  ap_time_indices, a_view_indices, p_view_indices = get_tcn_anchor_pos_indices(
      seq_len, num_views, num_pairs, window)

  # Gather the image strings.
  combined_anchor_indices = tf.concat(
      [tf.expand_dims(a_view_indices, 1),
       tf.expand_dims(ap_time_indices, 1)], 1)
  combined_pos_indices = tf.concat(
      [tf.expand_dims(p_view_indices, 1),
       tf.expand_dims(ap_time_indices, 1)], 1)
  anchor_images = tf.gather_nd(views, combined_anchor_indices)
  pos_images = tf.gather_nd(views, combined_pos_indices)

  # Decode images.
  anchor_images = tf.map_fn(
      preprocessing.decode_image, anchor_images, dtype=tf.float32)
  pos_images = tf.map_fn(
      preprocessing.decode_image, pos_images, dtype=tf.float32)

  # Concatenate [anchor, postitive] images into a batch and preprocess it.
  concatenated = tf.concat([anchor_images, pos_images], 0)
  preprocessed = preprocess_fn(concatenated, is_training)
  anchor_prepro, positive_prepro = tf.split(preprocessed, num_or_size_splits=2,
                                            axis=0)

  # Set static batch dimensions for all image tensors
  ims = [anchor_prepro, positive_prepro, anchor_images, pos_images]
  ims = [set_image_tensor_batch_dim(i, num_pairs) for i in ims]
  [anchor_prepro, positive_prepro, anchor_images, pos_images] = ims

  # Assign each anchor and positive the same label.
  anchor_labels = tf.range(1, num_pairs+1)
  positive_labels = tf.range(1, num_pairs+1)

  return (anchor_prepro, positive_prepro, anchor_images, pos_images,
          anchor_labels, positive_labels, seq_len)