Python functional_ops.map_fn方法代碼示例

# 需要導入模塊: from tensorflow.python.ops import functional_ops [as 別名]
# 或者: from tensorflow.python.ops.functional_ops import map_fn [as 別名]
def inception_logits(images=inception_images, num_splits=1):
    # images = tf.transpose(images, [0, 2, 3, 1])
    size = 299
    images = tf.image.resize_bilinear(images, [size, size])
    generated_images_list = array_ops.split(
        images, num_or_size_splits=num_splits)
    logits_ = functional_ops.map_fn(
        fn=functools.partial(tfgan.eval.run_inception, output_tensor='logits:0'),
        elems=array_ops.stack(generated_images_list),
        parallel_iterations=1,
        back_prop=False,
        swap_memory=True,
        name='RunClassifier')
    logits_ = array_ops.concat(array_ops.unstack(logits_), 0)

    return logits_ 

# 需要導入模塊: from tensorflow.python.ops import functional_ops [as 別名]
# 或者: from tensorflow.python.ops.functional_ops import map_fn [as 別名]
def tensors_to_item(self, keys_to_tensors):
    """See base class."""
    image_buffer = keys_to_tensors[self._image_key]
    image_format = keys_to_tensors[self._format_key]

    if self._repeated:
      return functional_ops.map_fn(lambda x: self._decode(x, image_format),
                                   image_buffer, dtype=self._dtype)
    else:
      return self._decode(image_buffer, image_format) 

# 需要導入模塊: from tensorflow.python.ops import functional_ops [as 別名]
# 或者: from tensorflow.python.ops.functional_ops import map_fn [as 別名]
def map_fn(fn, elems, name=None, dtype=None):
  """Map the function fn over the elements elems and return the outputs.

  Arguments:
      fn: Callable that will be called upon each element in elems
      elems: tensor
      name: A string name for the map node in the graph
      dtype: Output data type.

  Returns:
      Tensor with dtype `dtype`.
  """
  return functional_ops.map_fn(fn, elems, name=name, dtype=dtype) 

# 需要導入模塊: from tensorflow.python.ops import functional_ops [as 別名]
# 或者: from tensorflow.python.ops.functional_ops import map_fn [as 別名]
def sparse_boolean_mask(sparse_tensor, mask, name="sparse_boolean_mask"):
  """Boolean mask for `SparseTensor`s.

  Args:
    sparse_tensor: a `SparseTensor`.
    mask: a 1D boolean dense`Tensor` whose length is equal to the 0th dimension
      of `sparse_tensor`.
    name: optional name for this operation.
  Returns:
    A `SparseTensor` that contains row `k` of `sparse_tensor` iff `mask[k]` is
    `True`.
  """
  # TODO(jamieas): consider mask dimension > 1 for symmetry with `boolean_mask`.
  with ops.name_scope(name, values=[sparse_tensor, mask]):
    mask = ops.convert_to_tensor(mask)
    mask_rows = array_ops.where(mask)
    first_indices = array_ops.squeeze(array_ops.slice(sparse_tensor.indices,
                                                      [0, 0], [-1, 1]))

    # Identify indices corresponding to the rows identified by mask_rows.
    sparse_entry_matches = functional_ops.map_fn(
        lambda x: math_ops.equal(first_indices, x),
        mask_rows,
        dtype=dtypes.bool)
    # Combine the rows of index_matches to form a mask for the sparse indices
    # and values.
    to_retain = array_ops.reshape(
        functional_ops.foldl(math_ops.logical_or, sparse_entry_matches), [-1])

    return sparse_ops.sparse_retain(sparse_tensor, to_retain) 

# 需要導入模塊: from tensorflow.python.ops import functional_ops [as 別名]
# 或者: from tensorflow.python.ops.functional_ops import map_fn [as 別名]
def inception_logits(images, num_splits = 1):
    images = tf.transpose(images, [0, 2, 3, 1])
    size = 299
    images = tf.image.resize_bilinear(images, [size, size])
    generated_images_list = array_ops.split(images, num_or_size_splits = num_splits)
    logits = functional_ops.map_fn(
        fn = functools.partial(tfgan.eval.run_inception, output_tensor = 'logits:0'),
        elems = array_ops.stack(generated_images_list),
        parallel_iterations = 1,
        back_prop = False,
        swap_memory = True,
        name = 'RunClassifier')
    logits = array_ops.concat(array_ops.unstack(logits), 0)
    return logits 

# 需要導入模塊: from tensorflow.python.ops import functional_ops [as 別名]
# 或者: from tensorflow.python.ops.functional_ops import map_fn [as 別名]
def inception_activations(images, num_splits=1):
    images = tf.transpose(images, [0, 2, 3, 1])
    size = 299
    images = tf.image.resize_bilinear(images, [size, size])
    generated_images_list = array_ops.split(images, num_or_size_splits=num_splits)
    activations = functional_ops.map_fn(
        fn=functools.partial(tfgan.eval.run_inception, output_tensor='pool_3:0'),
        elems=array_ops.stack(generated_images_list),
        parallel_iterations=1,
        back_prop=False,
        swap_memory=True,
        name='RunClassifier')
    activations = array_ops.concat(array_ops.unstack(activations), 0)
    return activations 

# 需要導入模塊: from tensorflow.python.ops import functional_ops [as 別名]
# 或者: from tensorflow.python.ops.functional_ops import map_fn [as 別名]
def inception_logits(images=inception_images, num_splits=1):
    images=tf.transpose(images,[0,2,3,1])
    size = 299
    images = tf.image.resize_bilinear(images, [size, size])
    generated_images_list = array_ops.split(
    images, num_or_size_splits=num_splits)
    logits = functional_ops.map_fn(
        fn=functools.partial(tfgan.eval.run_inception, output_tensor='logits:0'),
        elems=array_ops.stack(generated_images_list),
        parallel_iterations=1,
        back_prop=False,
        swap_memory=True,
        name='RunClassifier')
    logits = array_ops.concat(array_ops.unstack(logits), 0)
    return logits 

# 需要導入模塊: from tensorflow.python.ops import functional_ops [as 別名]
# 或者: from tensorflow.python.ops.functional_ops import map_fn [as 別名]
def map_fn(fn, labeled_tensor, name=None):
  """Map on the list of tensors unpacked from labeled_tensor.

  See tf.map_fn.

  Args:
    fn: The function to apply to each unpacked LabeledTensor.
      It should have type LabeledTensor -> LabeledTensor.
    labeled_tensor: The input tensor.
    name: Optional op name.

  Returns:
    A tensor that packs the results of applying fn to the list of tensors
    unpacked from labeled_tensor.
  """
  with ops.name_scope(name, 'lt_map_fn', [labeled_tensor]) as scope:
    labeled_tensor = core.convert_to_labeled_tensor(labeled_tensor)

    unpack_lts = unpack(labeled_tensor)

    # TODO(ericmc): Fix this upstream.
    if labeled_tensor.dtype == dtypes.string:
      # We must construct the full graph here, because functional_ops.map_fn
      # doesn't work for string-valued tensors.
      # Constructing the full graph may be slow.
      map_lts = [fn(t) for t in unpack_lts]
      return pack(map_lts, list(labeled_tensor.axes.values())[0], name=scope)
    else:
      # Figure out what the axis labels should be, but use tf.map_fn to
      # construct the graph because it's efficient.
      # It may be slow to construct the full graph, so we infer the labels from
      # the first element.
      # TODO(ericmc): This builds a subgraph which then gets thrown away.
      # Find a more elegant solution.
      first_map_lt = fn(unpack_lts[0])
      final_axes = list(labeled_tensor.axes.values())[:1] + list(
          first_map_lt.axes.values())

      @tc.returns(ops.Tensor)
      @tc.accepts(ops.Tensor)
      def tf_fn(tensor):
        original_axes = list(labeled_tensor.axes.values())[1:]
        tensor_lt = core.LabeledTensor(tensor, original_axes)
        return fn(tensor_lt).tensor

      map_op = functional_ops.map_fn(tf_fn, labeled_tensor.tensor)
      map_lt = core.LabeledTensor(map_op, final_axes)

      return core.identity(map_lt, name=scope) 

# 需要導入模塊: from tensorflow.python.ops import functional_ops [as 別名]
# 或者: from tensorflow.python.ops.functional_ops import map_fn [as 別名]
def compress(self, inputs):
    """Compress inputs and store their binary representations into strings.

    Args:
      inputs: `Tensor` with values to be compressed.

    Returns:
      String `Tensor` vector containing the compressed representation of each
      batch element of `inputs`.
    """
    with ops.name_scope(self._name_scope()):
      inputs = ops.convert_to_tensor(inputs)
      if not self.built:
        # Check input assumptions set before layer building, e.g. input rank.
        input_spec.assert_input_compatibility(self.input_spec, inputs, self.name)
        if self.dtype is None:
          self._dtype = inputs.dtype.base_dtype.name
        self.build(inputs.shape)

      # Check input assumptions set after layer building, e.g. input shape.
      if not context.executing_eagerly():
        input_spec.assert_input_compatibility(self.input_spec, inputs, self.name)

      ndim = self.input_spec.ndim
      channel_axis = self._channel_axis(ndim)
      # Tuple of slices for expanding dimensions of tensors below.
      slices = ndim * [None] + [slice(None)]
      slices[channel_axis] = slice(None)
      slices = tuple(slices)

      # Expand dimensions of CDF to input dimensions, keeping the channels along
      # the right dimension.
      cdf = self._quantized_cdf[slices[1:]]
      num_levels = array_ops.shape(cdf)[-1] - 1

      # Bring inputs to the right range by centering the range on the medians.
      half = constant_op.constant(.5, dtype=self.dtype)
      medians = array_ops.squeeze(self._medians, [1, 2])
      offsets = (math_ops.cast(num_levels // 2, self.dtype) + half) - medians
      # Expand offsets to input dimensions and add to inputs.
      values = inputs + offsets[slices[:-1]]

      # Clip to range and cast to integers. Because we have added .5 above, and
      # all values are positive, the cast effectively implements rounding.
      values = math_ops.maximum(values, half)
      values = math_ops.minimum(
          values, math_ops.cast(num_levels, self.dtype) - half)
      values = math_ops.cast(values, dtypes.int16)

      def loop_body(tensor):
        return coder_ops.range_encode(
            tensor, cdf, precision=self.range_coder_precision)
      strings = functional_ops.map_fn(
          loop_body, values, dtype=dtypes.string, back_prop=False)

      if not context.executing_eagerly():
        strings.set_shape(inputs.shape[:1])

      return strings 

# 需要導入模塊: from tensorflow.python.ops import functional_ops [as 別名]
# 或者: from tensorflow.python.ops.functional_ops import map_fn [as 別名]
def inception_score(images, img_size=(299, 299), n_splits=10):
    """ referenced from https://github.com/tsc2017/Inception-Score/blob/master/inception_score.py """
    assert type(images) == np.ndarray
    assert len(images.shape) == 4
    assert images.shape[-1] == 3

    images = np.clip(images, 0., 255.)  # clipped into [0, 255]

    def inception_feat(img, n_splits=1):
        # img = tf.transpose(img, [0, 2, 3, 1])
        img = tf.image.resize_bilinear(img, img_size)

        generated_images_list = array_ops.split(img, num_or_size_splits=n_splits)

        logits = functional_ops.map_fn(
            fn=functools.partial(tf.contrib.gan.eval.run_inception, output_tensor="logits:0"),
            elems=array_ops.stack(generated_images_list),
            parallel_iterations=1,
            back_prop=False,
            swap_memory=True,
            name="RunClassifier"
        )
        logits = array_ops.concat(array_ops.unstack(logits), axis=0)
        return logits

    inception_images = tf.placeholder(tf.float32, [None, None, None, 3], name="inception-images")
    logits = inception_feat(inception_images)

    def get_inception_probs(x, n_classes=1000):
        n_batches = len(x) // batch_size

        preds = np.zeros([len(x), n_classes], dtype=np.float32)
        for i in range(n_batches):
            inp = x[i * batch_size:(i + 1) * batch_size] / 255. * 2 - 1.  # scaled into [-1, 1]
            preds[i * batch_size:(i + 1) * batch_size] = logits.eval({inception_images: inp})[:, :n_classes]
        preds = np.exp(preds) / np.sum(np.exp(preds), 1, keepdims=True)
        return preds

    def preds2score(preds, splits=10):
        scores = []
        for i in range(splits):
            part = preds[(i * preds.shape[0] // splits):((i + 1) * preds.shape[0] // splits), :]
            kl = part * (np.log(part) - np.log(np.expand_dims(np.mean(part, axis=0), axis=0)))
            kl = np.mean(np.sum(kl, axis=1))
            scores.append(np.exp(kl))
        return np.mean(scores), np.std(scores)

    preds = get_inception_probs(images)
    mean, std = preds2score(preds, splits=n_splits)
    return mean, std 

# 需要導入模塊: from tensorflow.python.ops import functional_ops [as 別名]
# 或者: from tensorflow.python.ops.functional_ops import map_fn [as 別名]
def fid_score(real_img, fake_img, img_size=(299, 299), n_splits=10):
    assert type(real_img) == np.ndarray and type(fake_img) == np.ndarray
    assert len(real_img.shape) == 4 and len(fake_img.shape) == 4
    assert real_img.shape[-1] == 3 and fake_img.shape[-1] == 3
    assert real_img.shape == fake_img.shape

    real_img = np.clip(real_img, 0., 255.)  # clipped into [0, 255]
    fake_img = np.clip(fake_img, 0., 255.)  # clipped into [0, 255]

    inception_images = tf.placeholder(tf.float32, [None, None, None, 3], name="inception-images")
    real_acts = tf.placeholder(tf.float32, [None, None], name="real_activations")
    fake_acts = tf.placeholder(tf.float32, [None, None], name="fake_activations")

    def inception_activation(images, n_splits=1):
        # images = tf.transpose(images, [0, 2, 3, 1])
        images = tf.image.resize_bilinear(images, img_size)

        generated_images_list = array_ops.split(images, num_or_size_splits=n_splits)

        acts = functional_ops.map_fn(
            fn=functools.partial(tf.contrib.gan.eval.run_inception, output_tensor="pool_3:0"),
            elems=array_ops.stack(generated_images_list),
            parallel_iterations=1,
            back_prop=False,
            swap_memory=True,
            name="RunClassifier"
        )
        acts = array_ops.concat(array_ops.unstack(acts), axis=0)
        return acts

    activations = inception_activation(inception_images)

    def get_inception_activations(x, feats=2048):
        n_batches = len(x) // batch_size

        acts = np.zeros([len(x), feats], dtype=np.float32)
        for i in range(n_batches):
            inp = x[i * batch_size:(i + 1) * batch_size] / 255. * 2 - 1.  # scaled into [-1, 1]
            acts[i * batch_size:(i + 1) * batch_size] = activations.eval({inception_images: inp})
        acts = np.exp(acts) / np.sum(np.exp(acts), 1, keepdims=True)
        return acts

    def get_fid(real, fake):
        return tf.contrib.gan.eval.frechet_classifier_distance_from_activations(real_acts, fake_acts).eval(
            feed_dict={
                real_acts: real,
                fake_acts: fake,
            }
        )

    real_img_acts = get_inception_activations(real_img)
    fake_img_acts = get_inception_activations(fake_img)

    fid = get_fid(real_img_acts, fake_img_acts)
    return fid