def test_filter_groundtruth_with_nan_box_coordinates(self):
input_tensors = {
fields.InputDataFields.groundtruth_boxes:
[[np.nan, np.nan, np.nan, np.nan], [0.2, 0.4, 0.1, 0.8]],
fields.InputDataFields.groundtruth_classes:
[1, 2],
fields.InputDataFields.groundtruth_is_crowd:
[False, True],
fields.InputDataFields.groundtruth_area:
[100.0, 238.7]
}
expected_tensors = {
fields.InputDataFields.groundtruth_boxes:
[[0.2, 0.4, 0.1, 0.8]],
fields.InputDataFields.groundtruth_classes:
[2],
fields.InputDataFields.groundtruth_is_crowd:
[True],
fields.InputDataFields.groundtruth_area:
[238.7]
}
output_tensors = ops.filter_groundtruth_with_nan_box_coordinates(
input_tensors)
with self.test_session() as sess:
output_tensors = sess.run(output_tensors)
for key in [fields.InputDataFields.groundtruth_boxes,
fields.InputDataFields.groundtruth_area]:
self.assertAllClose(expected_tensors[key], output_tensors[key])
for key in [fields.InputDataFields.groundtruth_classes,
fields.InputDataFields.groundtruth_is_crowd]:
self.assertAllEqual(expected_tensors[key], output_tensors[key])
def test_filter_groundtruth_with_nan_box_coordinates(self):
input_tensors = {
fields.InputDataFields.groundtruth_boxes:
[[np.nan, np.nan, np.nan, np.nan], [0.2, 0.4, 0.1, 0.8]],
fields.InputDataFields.groundtruth_classes:
[1, 2],
fields.InputDataFields.groundtruth_is_crowd:
[False, True],
fields.InputDataFields.groundtruth_area:
[100.0, 238.7]
}
expected_tensors = {
fields.InputDataFields.groundtruth_boxes:
[[0.2, 0.4, 0.1, 0.8]],
fields.InputDataFields.groundtruth_classes:
[2],
fields.InputDataFields.groundtruth_is_crowd:
[True],
fields.InputDataFields.groundtruth_area:
[238.7]
}
output_tensors = ops.filter_groundtruth_with_nan_box_coordinates(
input_tensors)
with self.test_session() as sess:
output_tensors = sess.run(output_tensors)
for key in [fields.InputDataFields.groundtruth_boxes,
fields.InputDataFields.groundtruth_area]:
self.assertAllClose(expected_tensors[key], output_tensors[key])
for key in [fields.InputDataFields.groundtruth_classes,
fields.InputDataFields.groundtruth_is_crowd]:
self.assertAllEqual(expected_tensors[key], output_tensors[key])
def _build_training_batch_dict(batch_sequences_with_states, unroll_length,
batch_size):
"""Builds training batch samples.
Args:
batch_sequences_with_states: A batch_sequences_with_states object.
unroll_length: Unrolled length for LSTM training.
batch_size: Batch size for queue outputs.
Returns:
A dictionary of tensors based on items in input_reader_config.
"""
seq_tensors_dict = {
fields.InputDataFields.image: [],
fields.InputDataFields.groundtruth_boxes: [],
fields.InputDataFields.groundtruth_classes: [],
'batch': batch_sequences_with_states,
}
for i in range(unroll_length):
for j in range(batch_size):
filtered_dict = util_ops.filter_groundtruth_with_nan_box_coordinates(
{
fields.InputDataFields.groundtruth_boxes:
(batch_sequences_with_states.sequences['groundtruth_boxes']
[j][i]),
fields.InputDataFields.groundtruth_classes:
(batch_sequences_with_states.
sequences['groundtruth_classes'][j][i]),
})
filtered_dict = util_ops.retain_groundtruth_with_positive_classes(
filtered_dict)
seq_tensors_dict[fields.InputDataFields.image].append(
batch_sequences_with_states.sequences['image'][j][i])
seq_tensors_dict[fields.InputDataFields.groundtruth_boxes].append(
filtered_dict[fields.InputDataFields.groundtruth_boxes])
seq_tensors_dict[
fields.InputDataFields.groundtruth_classes].append(
filtered_dict[fields.InputDataFields.groundtruth_classes])
seq_tensors_dict[fields.InputDataFields.image] = tuple(
seq_tensors_dict[fields.InputDataFields.image])
seq_tensors_dict[fields.InputDataFields.groundtruth_boxes] = tuple(
seq_tensors_dict[fields.InputDataFields.groundtruth_boxes])
seq_tensors_dict[fields.InputDataFields.groundtruth_classes] = tuple(
seq_tensors_dict[fields.InputDataFields.groundtruth_classes])
return seq_tensors_dict
def _build_training_batch_dict(batch_sequences_with_states, unroll_length,
batch_size):
"""Builds training batch samples.
Args:
batch_sequences_with_states: A batch_sequences_with_states object.
unroll_length: Unrolled length for LSTM training.
batch_size: Batch size for queue outputs.
Returns:
A dictionary of tensors based on items in input_reader_config.
"""
seq_tensors_dict = {
fields.InputDataFields.image: [],
fields.InputDataFields.groundtruth_boxes: [],
fields.InputDataFields.groundtruth_classes: [],
'batch': batch_sequences_with_states,
}
for i in range(unroll_length):
for j in range(batch_size):
filtered_dict = util_ops.filter_groundtruth_with_nan_box_coordinates({
fields.InputDataFields.groundtruth_boxes: (
batch_sequences_with_states.sequences['groundtruth_boxes'][j][i]),
fields.InputDataFields.groundtruth_classes: (
batch_sequences_with_states.sequences['groundtruth_classes'][j][i]
),
})
filtered_dict = util_ops.retain_groundtruth_with_positive_classes(
filtered_dict)
seq_tensors_dict[fields.InputDataFields.image].append(
batch_sequences_with_states.sequences['image'][j][i])
seq_tensors_dict[fields.InputDataFields.groundtruth_boxes].append(
filtered_dict[fields.InputDataFields.groundtruth_boxes])
seq_tensors_dict[fields.InputDataFields.groundtruth_classes].append(
filtered_dict[fields.InputDataFields.groundtruth_classes])
seq_tensors_dict[fields.InputDataFields.image] = tuple(
seq_tensors_dict[fields.InputDataFields.image])
seq_tensors_dict[fields.InputDataFields.groundtruth_boxes] = tuple(
seq_tensors_dict[fields.InputDataFields.groundtruth_boxes])
seq_tensors_dict[fields.InputDataFields.groundtruth_classes] = tuple(
seq_tensors_dict[fields.InputDataFields.groundtruth_classes])
return seq_tensors_dict
def transform_input_data(tensor_dict,
model_preprocess_fn,
image_resizer_fn,
num_classes,
data_augmentation_fn=None,
merge_multiple_boxes=False,
retain_original_image=False,
use_multiclass_scores=False,
use_bfloat16=False,
retain_original_image_additional_channels=False):
"""A single function that is responsible for all input data transformations.
Data transformation functions are applied in the following order.
1. If key fields.InputDataFields.image_additional_channels is present in
tensor_dict, the additional channels will be merged into
fields.InputDataFields.image.
2. data_augmentation_fn (optional): applied on tensor_dict.
3. model_preprocess_fn: applied only on image tensor in tensor_dict.
4. image_resizer_fn: applied on original image and instance mask tensor in
tensor_dict.
5. one_hot_encoding: applied to classes tensor in tensor_dict.
6. merge_multiple_boxes (optional): when groundtruth boxes are exactly the
same they can be merged into a single box with an associated k-hot class
label.
Args:
tensor_dict: dictionary containing input tensors keyed by
fields.InputDataFields.
model_preprocess_fn: model's preprocess function to apply on image tensor.
This function must take in a 4-D float tensor and return a 4-D preprocess
float tensor and a tensor containing the true image shape.
image_resizer_fn: image resizer function to apply on groundtruth instance
`masks. This function must take a 3-D float tensor of an image and a 3-D
tensor of instance masks and return a resized version of these along with
the true shapes.
num_classes: number of max classes to one-hot (or k-hot) encode the class
labels.
data_augmentation_fn: (optional) data augmentation function to apply on
input `tensor_dict`.
merge_multiple_boxes: (optional) whether to merge multiple groundtruth boxes
and classes for a given image if the boxes are exactly the same.
retain_original_image: (optional) whether to retain original image in the
output dictionary.
use_multiclass_scores: whether to use multiclass scores as class targets
instead of one-hot encoding of `groundtruth_classes`. When
this is True and multiclass_scores is empty, one-hot encoding of
`groundtruth_classes` is used as a fallback.
use_bfloat16: (optional) a bool, whether to use bfloat16 in training.
retain_original_image_additional_channels: (optional) Whether to retain
original image additional channels in the output dictionary.
Returns:
A dictionary keyed by fields.InputDataFields containing the tensors obtained
after applying all the transformations.
"""
out_tensor_dict = tensor_dict.copy()
if fields.InputDataFields.multiclass_scores in out_tensor_dict:
out_tensor_dict[
fields.InputDataFields
.multiclass_scores] = _multiclass_scores_or_one_hot_labels(
out_tensor_dict[fields.InputDataFields.multiclass_scores],
out_tensor_dict[fields.InputDataFields.groundtruth_boxes],
out_tensor_dict[fields.InputDataFields.groundtruth_classes],
num_classes)
if fields.InputDataFields.groundtruth_boxes in out_tensor_dict:
out_tensor_dict = util_ops.filter_groundtruth_with_nan_box_coordinates(
out_tensor_dict)
out_tensor_dict = util_ops.filter_unrecognized_classes(out_tensor_dict)
if retain_original_image:
out_tensor_dict[fields.InputDataFields.original_image] = tf.cast(
image_resizer_fn(out_tensor_dict[fields.InputDataFields.image],
None)[0], tf.uint8)
if fields.InputDataFields.image_additional_channels in out_tensor_dict:
channels = out_tensor_dict[fields.InputDataFields.image_additional_channels]
out_tensor_dict[fields.InputDataFields.image] = tf.concat(
[out_tensor_dict[fields.InputDataFields.image], channels], axis=2)
if retain_original_image_additional_channels:
out_tensor_dict[
fields.InputDataFields.image_additional_channels] = tf.cast(
image_resizer_fn(channels, None)[0], tf.uint8)
# Apply data augmentation ops.
if data_augmentation_fn is not None:
out_tensor_dict = data_augmentation_fn(out_tensor_dict)
# Apply model preprocessing ops and resize instance masks.
image = out_tensor_dict[fields.InputDataFields.image]
preprocessed_resized_image, true_image_shape = model_preprocess_fn(
tf.expand_dims(tf.cast(image, dtype=tf.float32), axis=0))
if use_bfloat16:
preprocessed_resized_image = tf.cast(
preprocessed_resized_image, tf.bfloat16)
out_tensor_dict[fields.InputDataFields.image] = tf.squeeze(
preprocessed_resized_image, axis=0)
out_tensor_dict[fields.InputDataFields.true_image_shape] = tf.squeeze(
true_image_shape, axis=0)
if fields.InputDataFields.groundtruth_instance_masks in out_tensor_dict:
masks = out_tensor_dict[fields.InputDataFields.groundtruth_instance_masks]
_, resized_masks, _ = image_resizer_fn(image, masks)
if use_bfloat16:
resized_masks = tf.cast(resized_masks, tf.bfloat16)
out_tensor_dict[
fields.InputDataFields.groundtruth_instance_masks] = resized_masks
label_offset = 1
zero_indexed_groundtruth_classes = out_tensor_dict[
fields.InputDataFields.groundtruth_classes] - label_offset
if use_multiclass_scores:
out_tensor_dict[
fields.InputDataFields.groundtruth_classes] = out_tensor_dict[
fields.InputDataFields.multiclass_scores]
else:
out_tensor_dict[fields.InputDataFields.groundtruth_classes] = tf.one_hot(
zero_indexed_groundtruth_classes, num_classes)
out_tensor_dict.pop(fields.InputDataFields.multiclass_scores, None)
if fields.InputDataFields.groundtruth_confidences in out_tensor_dict:
groundtruth_confidences = out_tensor_dict[
fields.InputDataFields.groundtruth_confidences]
# Map the confidences to the one-hot encoding of classes
out_tensor_dict[fields.InputDataFields.groundtruth_confidences] = (
tf.reshape(groundtruth_confidences, [-1, 1]) *
out_tensor_dict[fields.InputDataFields.groundtruth_classes])
else:
groundtruth_confidences = tf.ones_like(
zero_indexed_groundtruth_classes, dtype=tf.float32)
out_tensor_dict[fields.InputDataFields.groundtruth_confidences] = (
out_tensor_dict[fields.InputDataFields.groundtruth_classes])
if merge_multiple_boxes:
merged_boxes, merged_classes, merged_confidences, _ = (
util_ops.merge_boxes_with_multiple_labels(
out_tensor_dict[fields.InputDataFields.groundtruth_boxes],
zero_indexed_groundtruth_classes,
groundtruth_confidences,
num_classes))
merged_classes = tf.cast(merged_classes, tf.float32)
out_tensor_dict[fields.InputDataFields.groundtruth_boxes] = merged_boxes
out_tensor_dict[fields.InputDataFields.groundtruth_classes] = merged_classes
out_tensor_dict[fields.InputDataFields.groundtruth_confidences] = (
merged_confidences)
if fields.InputDataFields.groundtruth_boxes in out_tensor_dict:
out_tensor_dict[fields.InputDataFields.num_groundtruth_boxes] = tf.shape(
out_tensor_dict[fields.InputDataFields.groundtruth_boxes])[0]
return out_tensor_dict
def transform_input_data(tensor_dict,
model_preprocess_fn,
image_resizer_fn,
num_classes,
data_augmentation_fn=None,
merge_multiple_boxes=False,
retain_original_image=False):
"""A single function that is responsible for all input data transformations.
Data transformation functions are applied in the following order.
1. If key fields.InputDataFields.image_additional_channels is present in
tensor_dict, the additional channels will be merged into
fields.InputDataFields.image.
2. data_augmentation_fn (optional): applied on tensor_dict.
3. model_preprocess_fn: applied only on image tensor in tensor_dict.
4. image_resizer_fn: applied on original image and instance mask tensor in
tensor_dict.
5. one_hot_encoding: applied to classes tensor in tensor_dict.
6. merge_multiple_boxes (optional): when groundtruth boxes are exactly the
same they can be merged into a single box with an associated k-hot class
label.
Args:
tensor_dict: dictionary containing input tensors keyed by
fields.InputDataFields.
model_preprocess_fn: model's preprocess function to apply on image tensor.
This function must take in a 4-D float tensor and return a 4-D preprocess
float tensor and a tensor containing the true image shape.
image_resizer_fn: image resizer function to apply on groundtruth instance
`masks. This function must take a 3-D float tensor of an image and a 3-D
tensor of instance masks and return a resized version of these along with
the true shapes.
num_classes: number of max classes to one-hot (or k-hot) encode the class
data.
data_augmentation_fn: (optional) data augmentation function to apply on
input `tensor_dict`.
merge_multiple_boxes: (optional) whether to merge multiple groundtruth boxes
and classes for a given image if the boxes are exactly the same.
retain_original_image: (optional) whether to retain original image in the
output dictionary.
Returns:
A dictionary keyed by fields.InputDataFields containing the tensors obtained
after applying all the transformations.
"""
if fields.InputDataFields.groundtruth_boxes in tensor_dict:
tensor_dict = util_ops.filter_groundtruth_with_nan_box_coordinates(
tensor_dict)
if fields.InputDataFields.image_additional_channels in tensor_dict:
channels = tensor_dict[fields.InputDataFields.image_additional_channels]
tensor_dict[fields.InputDataFields.image] = tf.concat(
[tensor_dict[fields.InputDataFields.image], channels], axis=2)
if retain_original_image:
tensor_dict[fields.InputDataFields.original_image] = tf.cast(
tensor_dict[fields.InputDataFields.image], tf.uint8)
# Apply data augmentation ops.
if data_augmentation_fn is not None:
tensor_dict = data_augmentation_fn(tensor_dict)
# Apply model preprocessing ops and resize instance masks.
image = tensor_dict[fields.InputDataFields.image]
preprocessed_resized_image, true_image_shape = model_preprocess_fn(
tf.expand_dims(tf.to_float(image), axis=0))
tensor_dict[fields.InputDataFields.image] = tf.squeeze(
preprocessed_resized_image, axis=0)
tensor_dict[fields.InputDataFields.true_image_shape] = tf.squeeze(
true_image_shape, axis=0)
if fields.InputDataFields.groundtruth_instance_masks in tensor_dict:
masks = tensor_dict[fields.InputDataFields.groundtruth_instance_masks]
_, resized_masks, _ = image_resizer_fn(image, masks)
tensor_dict[fields.InputDataFields.
groundtruth_instance_masks] = resized_masks
# Transform groundtruth classes to one hot encodings.
label_offset = 1
zero_indexed_groundtruth_classes = tensor_dict[
fields.InputDataFields.groundtruth_classes] - label_offset
tensor_dict[fields.InputDataFields.groundtruth_classes] = tf.one_hot(
zero_indexed_groundtruth_classes, num_classes)
if merge_multiple_boxes:
merged_boxes, merged_classes, _ = util_ops.merge_boxes_with_multiple_labels(
tensor_dict[fields.InputDataFields.groundtruth_boxes],
zero_indexed_groundtruth_classes, num_classes)
merged_classes = tf.cast(merged_classes, tf.float32)
tensor_dict[fields.InputDataFields.groundtruth_boxes] = merged_boxes
tensor_dict[fields.InputDataFields.groundtruth_classes] = merged_classes
return tensor_dict
def transform_input_data(tensor_dict,
model_preprocess_fn,
image_resizer_fn,
num_classes,
data_augmentation_fn=None,
merge_multiple_boxes=False,
retain_original_image=False,
use_multiclass_scores=False):
"""A single function that is responsible for all input data transformations.
Data transformation functions are applied in the following order.
1. If key fields.InputDataFields.image_additional_channels is present in
tensor_dict, the additional channels will be merged into
fields.InputDataFields.image.
2. data_augmentation_fn (optional): applied on tensor_dict.
3. model_preprocess_fn: applied only on image tensor in tensor_dict.
4. image_resizer_fn: applied on original image and instance mask tensor in
tensor_dict.
5. one_hot_encoding: applied to classes tensor in tensor_dict.
6. merge_multiple_boxes (optional): when groundtruth boxes are exactly the
same they can be merged into a single box with an associated k-hot class
label.
Args:
tensor_dict: dictionary containing input tensors keyed by
fields.InputDataFields.
model_preprocess_fn: model's preprocess function to apply on image tensor.
This function must take in a 4-D float tensor and return a 4-D preprocess
float tensor and a tensor containing the true image shape.
image_resizer_fn: image resizer function to apply on groundtruth instance
`masks. This function must take a 3-D float tensor of an image and a 3-D
tensor of instance masks and return a resized version of these along with
the true shapes.
num_classes: number of max classes to one-hot (or k-hot) encode the class
labels.
data_augmentation_fn: (optional) data augmentation function to apply on
input `tensor_dict`.
merge_multiple_boxes: (optional) whether to merge multiple groundtruth boxes
and classes for a given image if the boxes are exactly the same.
retain_original_image: (optional) whether to retain original image in the
output dictionary.
use_multiclass_scores: whether to use multiclass scores as
class targets instead of one-hot encoding of `groundtruth_classes`.
use_bfloat16: (optional) a bool, whether to use bfloat16 in training.
Returns:
A dictionary keyed by fields.InputDataFields containing the tensors obtained
after applying all the transformations.
"""
# Reshape flattened multiclass scores tensor into a 2D tensor of shape
# [num_boxes, num_classes].
if fields.InputDataFields.multiclass_scores in tensor_dict:
tensor_dict[fields.InputDataFields.multiclass_scores] = tf.reshape(
tensor_dict[fields.InputDataFields.multiclass_scores], [
tf.shape(tensor_dict[fields.InputDataFields.groundtruth_boxes])[0],
num_classes
])
if fields.InputDataFields.groundtruth_boxes in tensor_dict:
tensor_dict = util_ops.filter_groundtruth_with_nan_box_coordinates(
tensor_dict)
tensor_dict = util_ops.filter_unrecognized_classes(tensor_dict)
if retain_original_image:
tensor_dict[fields.InputDataFields.original_image] = tf.cast(
image_resizer_fn(tensor_dict[fields.InputDataFields.image])[0],
tf.uint8)
if fields.InputDataFields.image_additional_channels in tensor_dict:
channels = tensor_dict[fields.InputDataFields.image_additional_channels]
tensor_dict[fields.InputDataFields.image] = tf.concat(
[tensor_dict[fields.InputDataFields.image], channels], axis=2)
# # Create gt_boxes_masks
# height, width, _ = tf.unstack(tf.shape(tensor_dict[fields.InputDataFields.image]))
# # image_template = tf.squeeze(tensor_dict[fields.InputDataFields.groundtruth_bel_O], axis=2)
# # image_template = tensor_dict[fields.InputDataFields.groundtruth_bel_O]
# label_boxes_list = tensor_dict[fields.InputDataFields.groundtruth_boxes]
# print('~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~')
# # print(image_template)
# print(label_boxes_list)
# boxes_mask = boxes2mask(label_boxes_list)
# tensor_dict[fields.InputDataFields.groundtruth_boxes_mask] = tf.stop_gradient(boxes_mask)
# # Create detection masks
# det_mask = tf.squeeze(tensor_dict[fields.InputDataFields.groundtruth_bel_O], axis=2)
# zeros = tf.zeros_like(det_mask)
# ones = tf.ones_like(det_mask)
# tensor_dict[fields.InputDataFields.groundtruth_boxes_mask] = tf.stop_gradient(tf.where(det_mask > 0, ones, zeros))
# Apply data augmentation ops.
if data_augmentation_fn is not None:
tensor_dict = data_augmentation_fn(tensor_dict) # todo first without data augm
# Apply model preprocessing ops and resize instance masks.
image = tensor_dict[fields.InputDataFields.image]
preprocessed_resized_image, true_image_shape = model_preprocess_fn(
tf.expand_dims(tf.cast(image, dtype=tf.float32), axis=0))
tensor_dict[fields.InputDataFields.image] = tf.squeeze(
preprocessed_resized_image, axis=0)
tensor_dict[fields.InputDataFields.true_image_shape] = tf.squeeze(
true_image_shape, axis=0)
groundtruth_bel_F = tensor_dict[fields.InputDataFields.groundtruth_bel_F]
groundtruth_bel_O = tensor_dict[fields.InputDataFields.groundtruth_bel_O]
groundtruth_z_max_detections = tensor_dict[fields.InputDataFields.groundtruth_z_max_detections]
groundtruth_z_min_observations = tensor_dict[fields.InputDataFields.groundtruth_z_min_observations]
groundtruth_bel_U = tensor_dict[fields.InputDataFields.groundtruth_bel_U]
groundtruth_z_min_detections = tensor_dict[fields.InputDataFields.groundtruth_z_min_detections]
groundtruth_detections_drivingCorridor = tensor_dict[fields.InputDataFields.groundtruth_detections_drivingCorridor]
groundtruth_intensity = tensor_dict[fields.InputDataFields.groundtruth_intensity]
groundtruth_bel_F = tf.expand_dims(tf.squeeze(groundtruth_bel_F, axis=2), axis=0)
_, resized_groundtruth_bel_F, _ = image_resizer_fn(image, groundtruth_bel_F)
# resized_groundtruth_bel_F = image_resizer_fn(groundtruth_bel_F)
groundtruth_bel_O = tf.expand_dims(tf.squeeze(groundtruth_bel_O, axis=2), axis=0)
_, resized_groundtruth_bel_O, _ = image_resizer_fn(image, groundtruth_bel_O)
# resized_groundtruth_bel_O = image_resizer_fn(groundtruth_bel_O)
groundtruth_z_max_detections = tf.expand_dims(tf.squeeze(groundtruth_z_max_detections, axis=2), axis=0)
_, resized_groundtruth_z_max_detections, _ = image_resizer_fn(image, groundtruth_z_max_detections)
groundtruth_z_min_observations = tf.expand_dims(tf.squeeze(groundtruth_z_min_observations, axis=2), axis=0)
_, resized_groundtruth_z_min_observations, _ = image_resizer_fn(image, groundtruth_z_min_observations)
groundtruth_bel_U = tf.expand_dims(tf.squeeze(groundtruth_bel_U, axis=2), axis=0)
_, resized_groundtruth_bel_U, _ = image_resizer_fn(image, groundtruth_bel_U)
groundtruth_z_min_detections = tf.expand_dims(tf.squeeze(groundtruth_z_min_detections, axis=2), axis=0)
_, resized_groundtruth_z_min_detections, _ = image_resizer_fn(image, groundtruth_z_min_detections)
groundtruth_detections_drivingCorridor = tf.expand_dims(tf.squeeze(groundtruth_detections_drivingCorridor, axis=2), axis=0)
_, resized_groundtruth_detections_drivingCorridor, _ = image_resizer_fn(image, groundtruth_detections_drivingCorridor)
groundtruth_intensity = tf.expand_dims(tf.squeeze(groundtruth_intensity, axis=2), axis=0)
_, resized_groundtruth_intensity, _ = image_resizer_fn(image, groundtruth_intensity)
tensor_dict[fields.InputDataFields.groundtruth_bel_F] = tf.expand_dims(tf.squeeze(
resized_groundtruth_bel_F, axis=0), axis=2)
tensor_dict[fields.InputDataFields.groundtruth_bel_O] = tf.expand_dims(tf.squeeze(
resized_groundtruth_bel_O, axis=0), axis=2)
tensor_dict[fields.InputDataFields.groundtruth_z_min_observations] = tf.expand_dims(tf.squeeze(
resized_groundtruth_z_min_observations, axis=0), axis=2)
tensor_dict[fields.InputDataFields.groundtruth_z_max_detections] = tf.expand_dims(tf.squeeze(
resized_groundtruth_z_max_detections, axis=0), axis=2)
tensor_dict[fields.InputDataFields.groundtruth_bel_U] = tf.expand_dims(tf.squeeze(
resized_groundtruth_bel_U, axis=0), axis=2)
tensor_dict[fields.InputDataFields.groundtruth_detections_drivingCorridor] = tf.expand_dims(tf.squeeze(
resized_groundtruth_detections_drivingCorridor, axis=0), axis=2)
tensor_dict[fields.InputDataFields.groundtruth_z_min_detections] = tf.expand_dims(tf.squeeze(
resized_groundtruth_z_min_detections, axis=0), axis=2)
tensor_dict[fields.InputDataFields.groundtruth_intensity] = tf.expand_dims(tf.squeeze(
resized_groundtruth_intensity, axis=0), axis=2)
# Transform groundtruth classes to one hot encodings.
label_offset = 1
zero_indexed_groundtruth_classes = tensor_dict[
fields.InputDataFields.groundtruth_classes] - label_offset
tensor_dict[fields.InputDataFields.groundtruth_classes] = tf.one_hot(
zero_indexed_groundtruth_classes, num_classes)
if use_multiclass_scores:
tensor_dict[fields.InputDataFields.groundtruth_classes] = tensor_dict[
fields.InputDataFields.multiclass_scores]
tensor_dict.pop(fields.InputDataFields.multiclass_scores, None)
if fields.InputDataFields.groundtruth_confidences in tensor_dict:
groundtruth_confidences = tensor_dict[
fields.InputDataFields.groundtruth_confidences]
# Map the confidences to the one-hot encoding of classes
tensor_dict[fields.InputDataFields.groundtruth_confidences] = (
tf.reshape(groundtruth_confidences, [-1, 1]) *
tensor_dict[fields.InputDataFields.groundtruth_classes])
else:
groundtruth_confidences = tf.ones_like(
zero_indexed_groundtruth_classes, dtype=tf.float32)
tensor_dict[fields.InputDataFields.groundtruth_confidences] = (
tensor_dict[fields.InputDataFields.groundtruth_classes])
if merge_multiple_boxes:
merged_boxes, merged_classes, merged_confidences, _ = (
util_ops.merge_boxes_with_multiple_labels(
tensor_dict[fields.InputDataFields.groundtruth_boxes],
zero_indexed_groundtruth_classes,
groundtruth_confidences,
num_classes))
merged_classes = tf.cast(merged_classes, tf.float32)
tensor_dict[fields.InputDataFields.groundtruth_boxes] = merged_boxes
tensor_dict[fields.InputDataFields.groundtruth_classes] = merged_classes
tensor_dict[fields.InputDataFields.groundtruth_confidences] = (
merged_confidences)
if fields.InputDataFields.groundtruth_boxes in tensor_dict:
tensor_dict[fields.InputDataFields.num_groundtruth_boxes] = tf.shape(
tensor_dict[fields.InputDataFields.groundtruth_boxes])[0]
# if fields.InputDataFields.groundtruth_bel_F in tensor_dict:
# channels = tensor_dict[fields.InputDataFields.groundtruth_bel_F]
# tensor_dict[fields.InputDataFields.groundtruth_bel_F] = tf.concat(
# [tensor_dict[fields.InputDataFields.groundtruth_bel_F], channels], axis=2)
# """ValueError: Can't concatenate scalars (use tf.stack instead) for 'concat_10' (op: 'ConcatV2') with input shapes: [], [], []."""
# if fields.InputDataFields.groundtruth_bel_O in tensor_dict:
# channels = tensor_dict[fields.InputDataFields.groundtruth_bel_O]
# tensor_dict[fields.InputDataFields.groundtruth_bel_O] = tf.concat(
# [tensor_dict[fields.InputDataFields.groundtruth_bel_O], channels], axis=2)
# """ValueError: Can't concatenate scalars (use tf.stack instead) for 'concat_10' (op: 'ConcatV2') with input shapes: [], [], []."""
return tensor_dict
def transform_input_data(tensor_dict,
model_preprocess_fn,
image_resizer_fn,
num_classes,
data_augmentation_fn=None,
merge_multiple_boxes=False,
retain_original_image=False,
use_bfloat16=False):
"""A single function that is responsible for all input data transformations.
Data transformation functions are applied in the following order.
1. If key fields.InputDataFields.image_additional_channels is present in
tensor_dict, the additional channels will be merged into
fields.InputDataFields.image.
2. data_augmentation_fn (optional): applied on tensor_dict.
3. model_preprocess_fn: applied only on image tensor in tensor_dict.
4. image_resizer_fn: applied on original image and instance mask tensor in
tensor_dict.
5. one_hot_encoding: applied to classes tensor in tensor_dict.
6. merge_multiple_boxes (optional): when groundtruth boxes are exactly the
same they can be merged into a single box with an associated k-hot class
label.
Args:
tensor_dict: dictionary containing input tensors keyed by
fields.InputDataFields.
model_preprocess_fn: model's preprocess function to apply on image tensor.
This function must take in a 4-D float tensor and return a 4-D preprocess
float tensor and a tensor containing the true image shape.
image_resizer_fn: image resizer function to apply on groundtruth instance
`masks. This function must take a 3-D float tensor of an image and a 3-D
tensor of instance masks and return a resized version of these along with
the true shapes.
num_classes: number of max classes to one-hot (or k-hot) encode the class
labels.
data_augmentation_fn: (optional) data augmentation function to apply on
input `tensor_dict`.
merge_multiple_boxes: (optional) whether to merge multiple groundtruth boxes
and classes for a given image if the boxes are exactly the same.
retain_original_image: (optional) whether to retain original image in the
output dictionary.
use_bfloat16: (optional) a bool, whether to use bfloat16 in training.
Returns:
A dictionary keyed by fields.InputDataFields containing the tensors obtained
after applying all the transformations.
"""
if fields.InputDataFields.groundtruth_boxes in tensor_dict:
tensor_dict = util_ops.filter_groundtruth_with_nan_box_coordinates(
tensor_dict)
if fields.InputDataFields.image_additional_channels in tensor_dict:
channels = tensor_dict[fields.InputDataFields.image_additional_channels]
tensor_dict[fields.InputDataFields.image] = tf.concat(
[tensor_dict[fields.InputDataFields.image], channels], axis=2)
if retain_original_image:
tensor_dict[fields.InputDataFields.original_image] = tf.cast(
image_resizer_fn(tensor_dict[fields.InputDataFields.image], None)[0],
tf.uint8)
# Apply data augmentation ops.
if data_augmentation_fn is not None:
tensor_dict = data_augmentation_fn(tensor_dict)
# Apply model preprocessing ops and resize instance masks.
image = tensor_dict[fields.InputDataFields.image]
preprocessed_resized_image, true_image_shape = model_preprocess_fn(
tf.expand_dims(tf.to_float(image), axis=0))
if use_bfloat16:
preprocessed_resized_image = tf.cast(
preprocessed_resized_image, tf.bfloat16)
tensor_dict[fields.InputDataFields.image] = tf.squeeze(
preprocessed_resized_image, axis=0)
tensor_dict[fields.InputDataFields.true_image_shape] = tf.squeeze(
true_image_shape, axis=0)
if fields.InputDataFields.groundtruth_instance_masks in tensor_dict:
masks = tensor_dict[fields.InputDataFields.groundtruth_instance_masks]
_, resized_masks, _ = image_resizer_fn(image, masks)
if use_bfloat16:
resized_masks = tf.cast(resized_masks, tf.bfloat16)
tensor_dict[fields.InputDataFields.
groundtruth_instance_masks] = resized_masks
# Transform groundtruth classes to one hot encodings.
label_offset = 1
zero_indexed_groundtruth_classes = tensor_dict[
fields.InputDataFields.groundtruth_classes] - label_offset
tensor_dict[fields.InputDataFields.groundtruth_classes] = tf.one_hot(
zero_indexed_groundtruth_classes, num_classes)
if fields.InputDataFields.groundtruth_confidences in tensor_dict:
groundtruth_confidences = tensor_dict[
fields.InputDataFields.groundtruth_confidences]
tensor_dict[fields.InputDataFields.groundtruth_confidences] = (
tf.sparse_to_dense(
zero_indexed_groundtruth_classes,
[num_classes],
groundtruth_confidences,
validate_indices=False))
else:
groundtruth_confidences = tf.ones_like(
zero_indexed_groundtruth_classes, dtype=tf.float32)
tensor_dict[fields.InputDataFields.groundtruth_confidences] = (
tensor_dict[fields.InputDataFields.groundtruth_classes])
if merge_multiple_boxes:
merged_boxes, merged_classes, merged_confidences, _ = (
util_ops.merge_boxes_with_multiple_labels(
tensor_dict[fields.InputDataFields.groundtruth_boxes],
zero_indexed_groundtruth_classes,
groundtruth_confidences,
num_classes))
merged_classes = tf.cast(merged_classes, tf.float32)
tensor_dict[fields.InputDataFields.groundtruth_boxes] = merged_boxes
tensor_dict[fields.InputDataFields.groundtruth_classes] = merged_classes
tensor_dict[fields.InputDataFields.groundtruth_confidences] = (
merged_confidences)
if fields.InputDataFields.groundtruth_boxes in tensor_dict:
tensor_dict[fields.InputDataFields.num_groundtruth_boxes] = tf.shape(
tensor_dict[fields.InputDataFields.groundtruth_boxes])[0]
return tensor_dict
def transform_input_data(tensor_dict,
model_preprocess_fn,
image_resizer_fn,
num_classes,
data_augmentation_fn=None,
merge_multiple_boxes=False,
retain_original_image=False,
use_multiclass_scores=False,
use_bfloat16=False,
retain_original_image_additional_channels=False,
keypoint_type_weight=None):
"""A single function that is responsible for all input data transformations.
Data transformation functions are applied in the following order.
1. If key fields.InputDataFields.image_additional_channels is present in
tensor_dict, the additional channels will be merged into
fields.InputDataFields.image.
2. data_augmentation_fn (optional): applied on tensor_dict.
3. model_preprocess_fn: applied only on image tensor in tensor_dict.
4. keypoint_type_weight (optional): If groundtruth keypoints are in
the tensor dictionary, per-keypoint weights are produced. These weights are
initialized by `keypoint_type_weight` (or ones if left None).
Then, for all keypoints that are not visible, the weights are set to 0 (to
avoid penalizing the model in a loss function).
5. image_resizer_fn: applied on original image and instance mask tensor in
tensor_dict.
6. one_hot_encoding: applied to classes tensor in tensor_dict.
7. merge_multiple_boxes (optional): when groundtruth boxes are exactly the
same they can be merged into a single box with an associated k-hot class
label.
Args:
tensor_dict: dictionary containing input tensors keyed by
fields.InputDataFields.
model_preprocess_fn: model's preprocess function to apply on image tensor.
This function must take in a 4-D float tensor and return a 4-D preprocess
float tensor and a tensor containing the true image shape.
image_resizer_fn: image resizer function to apply on groundtruth instance
`masks. This function must take a 3-D float tensor of an image and a 3-D
tensor of instance masks and return a resized version of these along with
the true shapes.
num_classes: number of max classes to one-hot (or k-hot) encode the class
labels.
data_augmentation_fn: (optional) data augmentation function to apply on
input `tensor_dict`.
merge_multiple_boxes: (optional) whether to merge multiple groundtruth boxes
and classes for a given image if the boxes are exactly the same.
retain_original_image: (optional) whether to retain original image in the
output dictionary.
use_multiclass_scores: whether to use multiclass scores as class targets
instead of one-hot encoding of `groundtruth_classes`. When
this is True and multiclass_scores is empty, one-hot encoding of
`groundtruth_classes` is used as a fallback.
use_bfloat16: (optional) a bool, whether to use bfloat16 in training.
retain_original_image_additional_channels: (optional) Whether to retain
original image additional channels in the output dictionary.
keypoint_type_weight: A list (of length num_keypoints) containing
groundtruth loss weights to use for each keypoint. If None, will use a
weight of 1.
Returns:
A dictionary keyed by fields.InputDataFields containing the tensors obtained
after applying all the transformations.
"""
out_tensor_dict = tensor_dict.copy()
if fields.InputDataFields.multiclass_scores in out_tensor_dict:
out_tensor_dict[
fields.InputDataFields.
multiclass_scores] = _multiclass_scores_or_one_hot_labels(
out_tensor_dict[fields.InputDataFields.multiclass_scores],
out_tensor_dict[fields.InputDataFields.groundtruth_boxes],
out_tensor_dict[fields.InputDataFields.groundtruth_classes],
num_classes)
if fields.InputDataFields.groundtruth_boxes in out_tensor_dict:
out_tensor_dict = util_ops.filter_groundtruth_with_nan_box_coordinates(
out_tensor_dict)
out_tensor_dict = util_ops.filter_unrecognized_classes(out_tensor_dict)
if retain_original_image:
out_tensor_dict[fields.InputDataFields.original_image] = tf.cast(
image_resizer_fn(out_tensor_dict[fields.InputDataFields.image],
None)[0], tf.uint8)
if fields.InputDataFields.image_additional_channels in out_tensor_dict:
channels = out_tensor_dict[
fields.InputDataFields.image_additional_channels]
out_tensor_dict[fields.InputDataFields.image] = tf.concat(
[out_tensor_dict[fields.InputDataFields.image], channels], axis=2)
if retain_original_image_additional_channels:
out_tensor_dict[
fields.InputDataFields.image_additional_channels] = tf.cast(
image_resizer_fn(channels, None)[0], tf.uint8)
# Apply data augmentation ops.
if data_augmentation_fn is not None:
out_tensor_dict = data_augmentation_fn(out_tensor_dict)
# Apply model preprocessing ops and resize instance masks.
image = out_tensor_dict[fields.InputDataFields.image]
preprocessed_resized_image, true_image_shape = model_preprocess_fn(
tf.expand_dims(tf.cast(image, dtype=tf.float32), axis=0))
preprocessed_shape = tf.shape(preprocessed_resized_image)
new_height, new_width = preprocessed_shape[1], preprocessed_shape[2]
im_box = tf.stack([
0.0, 0.0,
tf.to_float(new_height) / tf.to_float(true_image_shape[0, 0]),
tf.to_float(new_width) / tf.to_float(true_image_shape[0, 1])
])
if fields.InputDataFields.groundtruth_boxes in tensor_dict:
bboxes = out_tensor_dict[fields.InputDataFields.groundtruth_boxes]
boxlist = box_list.BoxList(bboxes)
realigned_bboxes = box_list_ops.change_coordinate_frame(
boxlist, im_box)
out_tensor_dict[
fields.InputDataFields.groundtruth_boxes] = realigned_bboxes.get()
if fields.InputDataFields.groundtruth_keypoints in tensor_dict:
keypoints = out_tensor_dict[
fields.InputDataFields.groundtruth_keypoints]
realigned_keypoints = keypoint_ops.change_coordinate_frame(
keypoints, im_box)
out_tensor_dict[
fields.InputDataFields.groundtruth_keypoints] = realigned_keypoints
flds_gt_kpt = fields.InputDataFields.groundtruth_keypoints
flds_gt_kpt_vis = fields.InputDataFields.groundtruth_keypoint_visibilities
flds_gt_kpt_weights = fields.InputDataFields.groundtruth_keypoint_weights
if flds_gt_kpt_vis not in out_tensor_dict:
out_tensor_dict[flds_gt_kpt_vis] = tf.ones_like(
out_tensor_dict[flds_gt_kpt][:, :, 0], dtype=tf.bool)
out_tensor_dict[flds_gt_kpt_weights] = (
keypoint_ops.keypoint_weights_from_visibilities(
out_tensor_dict[flds_gt_kpt_vis], keypoint_type_weight))
if use_bfloat16:
preprocessed_resized_image = tf.cast(preprocessed_resized_image,
tf.bfloat16)
out_tensor_dict[fields.InputDataFields.image] = tf.squeeze(
preprocessed_resized_image, axis=0)
out_tensor_dict[fields.InputDataFields.true_image_shape] = tf.squeeze(
true_image_shape, axis=0)
if fields.InputDataFields.groundtruth_instance_masks in out_tensor_dict:
masks = out_tensor_dict[
fields.InputDataFields.groundtruth_instance_masks]
_, resized_masks, _ = image_resizer_fn(image, masks)
if use_bfloat16:
resized_masks = tf.cast(resized_masks, tf.bfloat16)
out_tensor_dict[
fields.InputDataFields.groundtruth_instance_masks] = resized_masks
label_offset = 1
zero_indexed_groundtruth_classes = out_tensor_dict[
fields.InputDataFields.groundtruth_classes] - label_offset
if use_multiclass_scores:
out_tensor_dict[
fields.InputDataFields.groundtruth_classes] = out_tensor_dict[
fields.InputDataFields.multiclass_scores]
else:
out_tensor_dict[
fields.InputDataFields.groundtruth_classes] = tf.one_hot(
zero_indexed_groundtruth_classes, num_classes)
out_tensor_dict.pop(fields.InputDataFields.multiclass_scores, None)
if fields.InputDataFields.groundtruth_confidences in out_tensor_dict:
groundtruth_confidences = out_tensor_dict[
fields.InputDataFields.groundtruth_confidences]
# Map the confidences to the one-hot encoding of classes
out_tensor_dict[fields.InputDataFields.groundtruth_confidences] = (
tf.reshape(groundtruth_confidences, [-1, 1]) *
out_tensor_dict[fields.InputDataFields.groundtruth_classes])
else:
groundtruth_confidences = tf.ones_like(
zero_indexed_groundtruth_classes, dtype=tf.float32)
out_tensor_dict[fields.InputDataFields.groundtruth_confidences] = (
out_tensor_dict[fields.InputDataFields.groundtruth_classes])
if merge_multiple_boxes:
merged_boxes, merged_classes, merged_confidences, _ = (
util_ops.merge_boxes_with_multiple_labels(
out_tensor_dict[fields.InputDataFields.groundtruth_boxes],
zero_indexed_groundtruth_classes, groundtruth_confidences,
num_classes))
merged_classes = tf.cast(merged_classes, tf.float32)
out_tensor_dict[
fields.InputDataFields.groundtruth_boxes] = merged_boxes
out_tensor_dict[
fields.InputDataFields.groundtruth_classes] = merged_classes
out_tensor_dict[fields.InputDataFields.groundtruth_confidences] = (
merged_confidences)
if fields.InputDataFields.groundtruth_boxes in out_tensor_dict:
out_tensor_dict[
fields.InputDataFields.num_groundtruth_boxes] = tf.shape(
out_tensor_dict[fields.InputDataFields.groundtruth_boxes])[0]
return out_tensor_dict
def transform_input_data(tensor_dict,
model_preprocess_fn,
image_resizer_fn,
num_classes,
data_augmentation_fn=None,
merge_multiple_boxes=False,
retain_original_image=False,
use_multiclass_scores=False,
use_bfloat16=False):
out_tensor_dict = tensor_dict.copy()
if fields.InputDataFields.multiclass_scores in out_tensor_dict:
out_tensor_dict[
fields.InputDataFields.
multiclass_scores] = _multiclass_scores_or_one_hot_labels(
out_tensor_dict[fields.InputDataFields.multiclass_scores],
out_tensor_dict[fields.InputDataFields.groundtruth_boxes],
out_tensor_dict[fields.InputDataFields.groundtruth_classes],
num_classes)
if fields.InputDataFields.groundtruth_boxes in out_tensor_dict:
out_tensor_dict = util_ops.filter_groundtruth_with_nan_box_coordinates(
out_tensor_dict)
out_tensor_dict = util_ops.filter_unrecognized_classes(out_tensor_dict)
if retain_original_image:
out_tensor_dict[fields.InputDataFields.original_image] = tf.cast(
image_resizer_fn(out_tensor_dict[fields.InputDataFields.image],
None)[0], tf.uint8)
if fields.InputDataFields.image_additional_channels in out_tensor_dict:
channels = out_tensor_dict[
fields.InputDataFields.image_additional_channels]
out_tensor_dict[fields.InputDataFields.image] = tf.concat(
[out_tensor_dict[fields.InputDataFields.image], channels], axis=2)
# Apply data augmentation ops.
if data_augmentation_fn is not None:
out_tensor_dict = data_augmentation_fn(out_tensor_dict)
# Apply model preprocessing ops and resize instance masks.
image = out_tensor_dict[fields.InputDataFields.image]
preprocessed_resized_image, true_image_shape = model_preprocess_fn(
tf.expand_dims(tf.cast(image, dtype=tf.float32), axis=0))
if use_bfloat16:
preprocessed_resized_image = tf.cast(preprocessed_resized_image,
tf.bfloat16)
out_tensor_dict[fields.InputDataFields.image] = tf.squeeze(
preprocessed_resized_image, axis=0)
out_tensor_dict[fields.InputDataFields.true_image_shape] = tf.squeeze(
true_image_shape, axis=0)
if fields.InputDataFields.groundtruth_instance_masks in out_tensor_dict:
masks = out_tensor_dict[
fields.InputDataFields.groundtruth_instance_masks]
_, resized_masks, _ = image_resizer_fn(image, masks)
if use_bfloat16:
resized_masks = tf.cast(resized_masks, tf.bfloat16)
out_tensor_dict[
fields.InputDataFields.groundtruth_instance_masks] = resized_masks
label_offset = 1
zero_indexed_groundtruth_classes = out_tensor_dict[
fields.InputDataFields.groundtruth_classes] - label_offset
if use_multiclass_scores:
out_tensor_dict[
fields.InputDataFields.groundtruth_classes] = out_tensor_dict[
fields.InputDataFields.multiclass_scores]
else:
out_tensor_dict[
fields.InputDataFields.groundtruth_classes] = tf.one_hot(
zero_indexed_groundtruth_classes, num_classes)
out_tensor_dict.pop(fields.InputDataFields.multiclass_scores, None)
if fields.InputDataFields.groundtruth_confidences in out_tensor_dict:
groundtruth_confidences = out_tensor_dict[
fields.InputDataFields.groundtruth_confidences]
# Map the confidences to the one-hot encoding of classes
out_tensor_dict[fields.InputDataFields.groundtruth_confidences] = (
tf.reshape(groundtruth_confidences, [-1, 1]) *
out_tensor_dict[fields.InputDataFields.groundtruth_classes])
else:
groundtruth_confidences = tf.ones_like(
zero_indexed_groundtruth_classes, dtype=tf.float32)
out_tensor_dict[fields.InputDataFields.groundtruth_confidences] = (
out_tensor_dict[fields.InputDataFields.groundtruth_classes])
if merge_multiple_boxes:
merged_boxes, merged_classes, merged_confidences, _ = (
util_ops.merge_boxes_with_multiple_labels(
out_tensor_dict[fields.InputDataFields.groundtruth_boxes],
zero_indexed_groundtruth_classes, groundtruth_confidences,
num_classes))
merged_classes = tf.cast(merged_classes, tf.float32)
out_tensor_dict[
fields.InputDataFields.groundtruth_boxes] = merged_boxes
out_tensor_dict[
fields.InputDataFields.groundtruth_classes] = merged_classes
out_tensor_dict[fields.InputDataFields.groundtruth_confidences] = (
merged_confidences)
if fields.InputDataFields.groundtruth_boxes in out_tensor_dict:
out_tensor_dict[
fields.InputDataFields.num_groundtruth_boxes] = tf.shape(
out_tensor_dict[fields.InputDataFields.groundtruth_boxes])[0]
return out_tensor_dict
