def last_predict_part(boxes_encodings, classes_predictions_with_background, feature_maps, preprocessed_inputs):
"""Predicts unpostprocessed tensors from input tensor.
This function takes an input batch of images and runs it through the forward
pass of the network to yield unpostprocessesed predictions.
A side effect of calling the predict method is that self._anchors is
populated with a box_list.BoxList of anchors. These anchors must be
constructed before the postprocess or loss functions can be called.
Args:
boxes_encodings:
classes_predictions_with_background:
feature_maps:
preprocessed_inputs: a [batch, height, width, channels] image tensor.
true_image_shapes: int32 tensor of shape [batch, 3] where each row is
of the form [height, width, channels] indicating the shapes
of true images in the resized images, as resized images can be padded
with zeros.
"""
anchor_generator = anchor_generator_builder.build('ssd_anchor_generator')
prediction_dict = post_processor(boxes_encodings, classes_predictions_with_background,
feature_maps, anchor_generator.num_anchors_per_location())
image_shape = shape_utils.combined_static_and_dynamic_shape(
preprocessed_inputs)
feature_map_spatial_dims = get_feature_map_spatial_dims(
feature_maps)
anchors = box_list_ops.concatenate(
anchor_generator.generate(
feature_map_spatial_dims,
im_height=image_shape[1],
im_width=image_shape[2]))
box_encodings = tf.concat(prediction_dict['box_encodings'], axis=1)
print("tf.concat box_encodings:", box_encodings.name)
if box_encodings.shape.ndims == 4 and box_encodings.shape[2] == 1:
box_encodings = tf.squeeze(box_encodings, axis=2)
class_predictions_with_background = tf.concat(
prediction_dict['class_predictions_with_background'], axis=1)
predictions_dict = {
'preprocessed_inputs': preprocessed_inputs,
'box_encodings': box_encodings,
'class_predictions_with_background':
class_predictions_with_background,
'feature_maps': feature_maps,
'anchors': anchors.get()
}
return predictions_dict, anchors
def _match_when_rows_are_empty():
"""Performs matching when the rows of similarity matrix are empty.
When the rows are empty, all detections are false positives. So we return
a tensor of -1's to indicate that the columns do not match to any rows.
Returns:
matches: int32 tensor indicating the row each column matches to.
"""
similarity_matrix_shape = shape_utils.combined_static_and_dynamic_shape(
similarity_matrix)
return -1 * tf.ones([similarity_matrix_shape[1]], dtype=tf.int32)
def get_feature_map_spatial_dims(feature_maps):
"""Return list of spatial dimensions for each feature map in a list.
Args:
feature_maps: a list of tensors where the ith tensor has shape
[batch, height_i, width_i, depth_i].
Returns:
a list of pairs (height, width) for each feature map in feature_maps
"""
feature_map_shapes = [
shape_utils.combined_static_and_dynamic_shape(
feature_map) for feature_map in feature_maps
]
return [(shape[1], shape[2]) for shape in feature_map_shapes]
def matmul_gather_on_zeroth_axis(params, indices, scope=None):
"""Matrix multiplication based implementation of tf.gather on zeroth axis.
TODO(rathodv, jonathanhuang): enable sparse matmul option.
Args:
params: A float32 Tensor. The tensor from which to gather values.
Must be at least rank 1.
indices: A Tensor. Must be one of the following types: int32, int64.
Must be in range [0, params.shape[0])
scope: A name for the operation (optional).
Returns:
A Tensor. Has the same type as params. Values from params gathered
from indices given by indices, with shape indices.shape + params.shape[1:].
"""
with tf.name_scope(scope, 'MatMulGather'):
params_shape = shape_utils.combined_static_and_dynamic_shape(params)
indices_shape = shape_utils.combined_static_and_dynamic_shape(indices)
params2d = tf.reshape(params, [params_shape[0], -1])
indicator_matrix = tf.one_hot(indices, params_shape[0])
gathered_result_flattened = tf.matmul(indicator_matrix, params2d)
return tf.reshape(gathered_result_flattened,
tf.stack(indices_shape + params_shape[1:]))
def _batch_decode_boxes(box_encodings, anchor_boxes):
"""Decodes box encodings with respect to the anchor boxes.
Args:
box_encodings: a 4-D tensor with shape
[batch_size, num_anchors, num_classes, self._box_coder.code_size]
representing box encodings.
anchor_boxes: [batch_size, num_anchors, self._box_coder.code_size]
representing decoded bounding boxes. If using a shared box across
classes the shape will instead be
[total_num_proposals, 1, self._box_coder.code_size].
Returns:
decoded_boxes: a
[batch_size, num_anchors, num_classes, self._box_coder.code_size]
float tensor representing bounding box predictions (for each image in
batch, proposal and class). If using a shared box across classes the
shape will instead be
[batch_size, num_anchors, 1, self._box_coder.code_size].
"""
combined_shape = shape_utils.combined_static_and_dynamic_shape(
box_encodings)
num_classes = combined_shape[2]
tiled_anchor_boxes = tf.tile(tf.expand_dims(anchor_boxes, 2),
[1, 1, num_classes, 1])
print("tiled_anchor_boxes:", tiled_anchor_boxes.name)
tiled_anchors_boxlist = box_list.BoxList(
tf.reshape(tiled_anchor_boxes, [-1, 4]))
_proposal_target_assigner = target_assigner.create_target_assigner(
'FasterRCNN', 'proposal')
_box_coder = _proposal_target_assigner.box_coder
decoded_boxes = _box_coder.decode(
tf.reshape(box_encodings, [-1, _box_coder.code_size]),
tiled_anchors_boxlist)
print("combined_shape[0]:", combined_shape[0])
print("combined_shape[1]:", combined_shape[1])
print("num_classes:", num_classes)
print("decoded_boxes.get():", decoded_boxes.get())
decoded_boxes_reahpe = tf.reshape(
decoded_boxes.get(),
tf.stack([combined_shape[0], combined_shape[1], num_classes, 4]))
return decoded_boxes_reahpe
def _match_when_rows_are_non_empty():
"""Performs matching when the rows of similarity matrix are non empty.
Returns:
matches: int32 tensor indicating the row each column matches to.
"""
# Matches for each column
matches = tf.argmax(similarity_matrix, 0, output_type=tf.int32)
# Deal with matched and unmatched threshold
if self._matched_threshold is not None:
# Get logical indices of ignored and unmatched columns as tf.int64
matched_vals = tf.reduce_max(similarity_matrix, 0)
below_unmatched_threshold = tf.greater(
self._unmatched_threshold, matched_vals)
between_thresholds = tf.logical_and(
tf.greater_equal(matched_vals, self._unmatched_threshold),
tf.greater(self._matched_threshold, matched_vals))
if self._negatives_lower_than_unmatched:
matches = self._set_values_using_indicator(
matches, below_unmatched_threshold, -1)
matches = self._set_values_using_indicator(
matches, between_thresholds, -2)
else:
matches = self._set_values_using_indicator(
matches, below_unmatched_threshold, -2)
matches = self._set_values_using_indicator(
matches, between_thresholds, -1)
if self._force_match_for_each_row:
similarity_matrix_shape = shape_utils.combined_static_and_dynamic_shape(
similarity_matrix)
force_match_column_ids = tf.argmax(similarity_matrix,
1,
output_type=tf.int32)
force_match_column_indicators = tf.one_hot(
force_match_column_ids, depth=similarity_matrix_shape[1])
force_match_row_ids = tf.argmax(force_match_column_indicators,
0,
output_type=tf.int32)
force_match_column_mask = tf.cast(
tf.reduce_max(force_match_column_indicators, 0), tf.bool)
final_matches = tf.where(force_match_column_mask,
force_match_row_ids, matches)
return final_matches
else:
return matches
def _flatten_first_two_dimensions(inputs):
"""Flattens `K-d` tensor along batch dimension to be a `(K-1)-d` tensor.
Converts `inputs` with shape [A, B, ..., depth] into a tensor of shape
[A * B, ..., depth].
Args:
inputs: A float tensor with shape [A, B, ..., depth]. Note that the first
two and last dimensions must be statically defined.
Returns:
A float tensor with shape [A * B, ..., depth] (where the first and last
dimension are statically defined.
"""
combined_shape = shape_utils.combined_static_and_dynamic_shape(inputs)
flattened_shape = tf.stack([combined_shape[0] * combined_shape[1]] +
combined_shape[2:])
return tf.reshape(inputs, flattened_shape)
def _batch_decode(anchors, box_encodings):
"""Decodes a batch of box encodings with respect to the anchors.
Args:
box_encodings: A float32 tensor of shape
[batch_size, num_anchors, box_code_size] containing box encodings.
Returns:
decoded_boxes: A float32 tensor of shape
[batch_size, num_anchors, 4] containing the decoded boxes.
decoded_keypoints: A float32 tensor of shape
[batch_size, num_anchors, num_keypoints, 2] containing the decoded
keypoints if present in the input `box_encodings`, None otherwise.
"""
combined_shape = shape_utils.combined_static_and_dynamic_shape(
box_encodings)
batch_size = combined_shape[0]
tiled_anchor_boxes = tf.tile(
tf.expand_dims(anchors.get(), 0), [batch_size, 1, 1])
tiled_anchors_boxlist = box_list.BoxList(
tf.reshape(tiled_anchor_boxes, [-1, 4]))
box_coder = box_coder_builder.build("faster_rcnn_box_coder")
decoded_boxes = box_coder.decode(
tf.reshape(box_encodings, [-1, box_coder.code_size]),
tiled_anchors_boxlist)
decoded_keypoints = None
if decoded_boxes.has_field(fields.BoxListFields.keypoints):
decoded_keypoints = decoded_boxes.get_field(
fields.BoxListFields.keypoints)
num_keypoints = decoded_keypoints.get_shape()[1]
decoded_keypoints = tf.reshape(
decoded_keypoints,
tf.stack([combined_shape[0], combined_shape[1], num_keypoints, 2]))
decoded_boxes = tf.reshape(decoded_boxes.get(), tf.stack(
[combined_shape[0], combined_shape[1], 4]))
return decoded_boxes, decoded_keypoints
def _compute_clip_window(preprocessed_images, true_image_shapes):
"""Computes clip window to use during post_processing.
Computes a new clip window to use during post-processing based on
`resized_image_shapes` and `true_image_shapes` only if `preprocess` method
has been called. Otherwise returns a default clip window of [0, 0, 1, 1].
Args:
preprocessed_images: the [batch, height, width, channels] image
tensor.
true_image_shapes: int32 tensor of shape [batch, 3] where each row is
of the form [height, width, channels] indicating the shapes
of true images in the resized images, as resized images can be padded
with zeros. Or None if the clip window should cover the full image.
Returns:
a 2-D float32 tensor of the form [batch_size, 4] containing the clip
window for each image in the batch in normalized coordinates (relative to
the resized dimensions) where each clip window is of the form [ymin, xmin,
ymax, xmax] or a default clip window of [0, 0, 1, 1].
"""
if true_image_shapes is None:
return tf.constant([0, 0, 1, 1], dtype=tf.float32)
resized_inputs_shape = shape_utils.combined_static_and_dynamic_shape(
preprocessed_images)
true_heights, true_widths, _ = tf.unstack(
tf.to_float(true_image_shapes), axis=1)
padded_height = tf.to_float(resized_inputs_shape[1])
padded_width = tf.to_float(resized_inputs_shape[2])
return tf.stack(
[
tf.zeros_like(true_heights),
tf.zeros_like(true_widths), true_heights / padded_height,
true_widths / padded_width
],
axis=1)
def _create_regression_targets(self, anchors, groundtruth_boxes, match):
"""Returns a regression target for each anchor.
Args:
anchors: a BoxList representing N anchors
groundtruth_boxes: a BoxList representing M groundtruth_boxes
match: a matcher.Match object
Returns:
reg_targets: a float32 tensor with shape [N, box_code_dimension]
"""
matched_gt_boxes = match.gather_based_on_match(
groundtruth_boxes.get(),
unmatched_value=tf.zeros(4),
ignored_value=tf.zeros(4))
matched_gt_boxlist = box_list.BoxList(matched_gt_boxes)
if groundtruth_boxes.has_field(fields.BoxListFields.keypoints):
groundtruth_keypoints = groundtruth_boxes.get_field(
fields.BoxListFields.keypoints)
matched_keypoints = match.gather_based_on_match(
groundtruth_keypoints,
unmatched_value=tf.zeros(
groundtruth_keypoints.get_shape()[1:]),
ignored_value=tf.zeros(groundtruth_keypoints.get_shape()[1:]))
matched_gt_boxlist.add_field(fields.BoxListFields.keypoints,
matched_keypoints)
matched_reg_targets = self._box_coder.encode(matched_gt_boxlist,
anchors)
match_results_shape = shape_utils.combined_static_and_dynamic_shape(
match.match_results)
# Zero out the unmatched and ignored regression targets.
unmatched_ignored_reg_targets = tf.tile(
self._default_regression_target(), [match_results_shape[0], 1])
matched_anchors_mask = match.matched_column_indicator()
reg_targets = tf.where(matched_anchors_mask, matched_reg_targets,
unmatched_ignored_reg_targets)
return reg_targets
def nearest_neighbor_upsampling(input_tensor, scale):
"""Nearest neighbor upsampling implementation.
Nearest neighbor upsampling function that maps input tensor with shape
[batch_size, height, width, channels] to [batch_size, height * scale
, width * scale, channels]. This implementation only uses reshape and
broadcasting to make it TPU compatible.
Args:
input_tensor: A float32 tensor of size [batch, height_in, width_in,
channels].
scale: An integer multiple to scale resolution of input data.
Returns:
data_up: A float32 tensor of size
[batch, height_in*scale, width_in*scale, channels].
"""
with tf.name_scope('nearest_neighbor_upsampling'):
(batch_size, height, width,
channels) = shape_utils.combined_static_and_dynamic_shape(input_tensor)
output_tensor = tf.reshape(
input_tensor, [batch_size, height, 1, width, 1, channels]) * tf.ones(
[1, 1, scale, 1, scale, 1], dtype=input_tensor.dtype)
return tf.reshape(output_tensor,
[batch_size, height * scale, width * scale, channels])
def _postprocess_rpn(rpn_box_encodings_batch,
rpn_objectness_predictions_with_background_batch, anchors,
image_shapes, first_stage_max_proposals):
"""Converts first stage prediction tensors from the RPN to proposals.
This function decodes the raw RPN predictions, runs non-max suppression
on the result.
Note that the behavior of this function is slightly modified during
training --- specifically, we stop the gradient from passing through the
proposal boxes and we only return a balanced sampled subset of proposals
with size `second_stage_batch_size`.
Args:
rpn_box_encodings_batch: A 3-D float32 tensor of shape
[batch_size, num_anchors, self._box_coder.code_size] containing
predicted proposal box encodings.
rpn_objectness_predictions_with_background_batch: A 3-D float tensor of
shape [batch_size, num_anchors, 2] containing objectness predictions
(logits) for each of the anchors with 0 corresponding to background
and 1 corresponding to object.
anchors: A 2-D tensor of shape [num_anchors, 4] representing anchors
for the first stage RPN. Note that `num_anchors` can differ depending
on whether the model is created in training or inference mode.
image_shapes: A 2-D tensor of shape [batch, 3] containing the shapes of
images in the batch.
true_image_shapes: int32 tensor of shape [batch, 3] where each row is
of the form [height, width, channels] indicating the shapes
of true images in the resized images, as resized images can be padded
with zeros.
Returns:
proposal_boxes: A float tensor with shape
[batch_size, max_num_proposals, 4] representing the (potentially zero
padded) proposal boxes for all images in the batch. These boxes are
represented as normalized coordinates.
proposal_scores: A float tensor with shape
[batch_size, max_num_proposals] representing the (potentially zero
padded) proposal objectness scores for all images in the batch.
num_proposals: A Tensor of type `int32`. A 1-D tensor of shape [batch]
representing the number of proposals predicted for each image in
the batch.
"""
first_stage_nms_score_threshold = 0.0
first_stage_nms_iou_threshold = 0.7
first_stage_max_proposals = first_stage_max_proposals
rpn_box_encodings_batch = tf.expand_dims(rpn_box_encodings_batch, axis=2)
print("rpn_box_encodings_batch name:", rpn_box_encodings_batch.name)
print("rpn_box_encodings_batch: shape", rpn_box_encodings_batch.shape)
rpn_encodings_shape = shape_utils.combined_static_and_dynamic_shape(
rpn_box_encodings_batch)
tiled_anchor_boxes = tf.tile(tf.expand_dims(anchors, 0),
[rpn_encodings_shape[0], 1, 1])
print("_batch_decode_boxes 1")
proposal_boxes = _batch_decode_boxes(rpn_box_encodings_batch,
tiled_anchor_boxes)
proposal_boxes = tf.squeeze(proposal_boxes, axis=2)
rpn_objectness_softmax_without_background = tf.nn.softmax(
rpn_objectness_predictions_with_background_batch)[:, :, 1]
clip_window = _compute_clip_window(image_shapes)
(proposal_boxes, proposal_scores, _, _, _,
num_proposals) = post_processing.batch_multiclass_non_max_suppression(
tf.expand_dims(proposal_boxes, axis=2),
tf.expand_dims(rpn_objectness_softmax_without_background, axis=2),
first_stage_nms_score_threshold,
first_stage_nms_iou_threshold,
first_stage_max_proposals,
first_stage_max_proposals,
clip_window=clip_window)
# normalize proposal boxes
def normalize_boxes(args):
proposal_boxes_per_image = args[0]
image_shape = args[1]
normalized_boxes_per_image = box_list_ops.to_normalized_coordinates(
box_list.BoxList(proposal_boxes_per_image),
image_shape[0],
image_shape[1],
check_range=False).get()
return normalized_boxes_per_image
normalized_proposal_boxes = shape_utils.static_or_dynamic_map_fn(
normalize_boxes,
elems=[proposal_boxes, image_shapes],
dtype=tf.float32)
return normalized_proposal_boxes, proposal_scores, num_proposals
def second_stage_box_predictor(preprocessed_inputs, box_encoding_reshape,
class_prediction_reshape, rpn_features_to_crop,
rpn_box_encodings,
rpn_objectness_predictions_with_background,
true_image_shapes, rpn_box_predictor_features):
image_shape = shape_utils.combined_static_and_dynamic_shape(
preprocessed_inputs)
first_stage_anchor_generator = anchor_generator_builder.build(
"grid_anchor_generator")
# The Faster R-CNN paper recommends pruning anchors that venture outside
# the image window at training time and clipping at inference time.
clip_window = tf.to_float(tf.stack([0, 0, image_shape[1], image_shape[2]]))
feature_map_shape = tf.shape(rpn_features_to_crop)
anchors_boxlist = box_list_ops.concatenate(
first_stage_anchor_generator.generate([(feature_map_shape[1],
feature_map_shape[2])]))
anchors_boxlist = box_list_ops.clip_to_window(anchors_boxlist, clip_window)
_anchors = anchors_boxlist
print("second_stage_box_predictor _postprocess_rpn")
image_shape_2d = _image_batch_shape_2d(image_shape)
num_anchors_per_location = (
first_stage_anchor_generator.num_anchors_per_location())
if len(num_anchors_per_location) != 1:
raise RuntimeError('anchor_generator is expected to generate anchors '
'corresponding to a single feature map.')
box_predictions = _first_stage_box_predictor_predict(
[rpn_box_predictor_features], [rpn_box_encodings],
[rpn_objectness_predictions_with_background], num_anchors_per_location)
predictions_box_encodings = tf.concat(box_predictions[BOX_ENCODINGS],
axis=1)
print("squeeze predictions_box_encodings.shape:",
predictions_box_encodings.shape)
rpn_box_encodings = tf.squeeze(predictions_box_encodings, axis=2)
print("rpn_box_encodings.shape:", rpn_box_encodings.shape)
rpn_objectness_predictions_with_background = tf.concat(
box_predictions[CLASS_PREDICTIONS_WITH_BACKGROUND], axis=1)
proposal_boxes_normalized, _, num_proposals = _postprocess_rpn(
rpn_box_encodings,
rpn_objectness_predictions_with_background,
_anchors.get(),
image_shape_2d,
first_stage_max_proposals=100)
prediction_dict = {
'rpn_box_predictor_features':
rpn_box_predictor_features,
'rpn_features_to_crop':
rpn_features_to_crop,
'image_shape':
image_shape,
'rpn_box_encodings':
rpn_box_encodings,
'rpn_objectness_predictions_with_background':
rpn_objectness_predictions_with_background,
}
refined_box_encodings = tf.squeeze(box_encoding_reshape,
axis=1,
name='all_refined_box_encodings')
class_predictions_with_background = tf.squeeze(
class_prediction_reshape,
axis=1,
name='all_class_predictions_with_background')
_parallel_iterations = 16
absolute_proposal_boxes = ops.normalized_to_image_coordinates(
proposal_boxes_normalized, image_shape, _parallel_iterations)
prediction_dict1 = {
'refined_box_encodings': refined_box_encodings,
'class_predictions_with_background': class_predictions_with_background,
'num_proposals': num_proposals,
'proposal_boxes': absolute_proposal_boxes,
}
prediction_dict.update(prediction_dict1)
result_output = second_postprocess(prediction_dict, true_image_shapes)
return result_output
def _first_stage_box_predictor_predict(image_features, box_encodings,
class_predictions_with_backgrounds,
num_predictions_per_locations):
"""Computes encoded object locations and corresponding confidences.
Args:
image_features: A list of float tensors of shape [batch_size, height_i,
width_i, channels_i] containing features for a batch of images.
num_predictions_per_location_list: A list of integers representing the
number of box predictions to be made per spatial location for each
feature map.
Returns:
box_encodings: A list of float tensors of shape
[batch_size, num_anchors_i, q, code_size] representing the location of
the objects, where q is 1 or the number of classes. Each entry in the
list corresponds to a feature map in the input `image_features` list.
class_predictions_with_background: A list of float tensors of shape
[batch_size, num_anchors_i, num_classes + 1] representing the class
predictions for the proposals. Each entry in the list corresponds to a
feature map in the input `image_features` list.
"""
box_encodings_list = []
class_predictions_list = []
num_classes = 1
num_class_slots = num_classes + 1
_proposal_target_assigner = target_assigner.create_target_assigner(
'FasterRCNN', 'proposal')
_box_coder = _proposal_target_assigner.box_coder
_box_code_size = _box_coder.code_size
for (image_feature, box_encoding, class_predictions_with_background,
num_predictions_per_location) in zip(
image_features, box_encodings, class_predictions_with_backgrounds,
num_predictions_per_locations):
combined_feature_map_shape = (
shape_utils.combined_static_and_dynamic_shape(image_feature))
print("_box_code_size:", _box_code_size)
print("num_predictions_per_location:", num_predictions_per_location)
print("combined_feature_map_shape[1]:", combined_feature_map_shape[1])
print("combined_feature_map_shape[2]:", combined_feature_map_shape[2])
print("box_encodings:", box_encoding.shape)
shapes = tf.stack([
combined_feature_map_shape[0], combined_feature_map_shape[1] *
combined_feature_map_shape[2] * num_predictions_per_location, 1,
_box_code_size
])
box_encoding_reshape = tf.reshape(box_encoding, shapes)
print("box_encoding_reshape:", box_encoding_reshape.shape)
box_encodings_list.append(box_encoding_reshape)
class_predictions_with_background = tf.reshape(
class_predictions_with_background,
tf.stack([
combined_feature_map_shape[0], combined_feature_map_shape[1] *
combined_feature_map_shape[2] * num_predictions_per_location,
num_class_slots
]))
class_predictions_list.append(class_predictions_with_background)
return {
BOX_ENCODINGS: box_encodings_list,
CLASS_PREDICTIONS_WITH_BACKGROUND: class_predictions_list
}
def assign(self,
anchors,
groundtruth_boxes,
groundtruth_labels=None,
groundtruth_weights=None,
**params):
"""Assign classification and regression targets to each anchor.
For a given set of anchors and groundtruth detections, match anchors
to groundtruth_boxes and assign classification and regression targets to
each anchor as well as weights based on the resulting match (specifying,
e.g., which anchors should not contribute to training loss).
Anchors that are not matched to anything are given a classification target
of self._unmatched_cls_target which can be specified via the constructor.
Args:
anchors: a BoxList representing N anchors
groundtruth_boxes: a BoxList representing M groundtruth boxes
groundtruth_labels: a tensor of shape [M, d_1, ... d_k]
with labels for each of the ground_truth boxes. The subshape
[d_1, ... d_k] can be empty (corresponding to scalar inputs). When set
to None, groundtruth_labels assumes a binary problem where all
ground_truth boxes get a positive label (of 1).
groundtruth_weights: a float tensor of shape [M] indicating the weight to
assign to all anchors match to a particular groundtruth box. The weights
must be in [0., 1.]. If None, all weights are set to 1.
**params: Additional keyword arguments for specific implementations of
the Matcher.
Returns:
cls_targets: a float32 tensor with shape [num_anchors, d_1, d_2 ... d_k],
where the subshape [d_1, ..., d_k] is compatible with groundtruth_labels
which has shape [num_gt_boxes, d_1, d_2, ... d_k].
cls_weights: a float32 tensor with shape [num_anchors]
reg_targets: a float32 tensor with shape [num_anchors, box_code_dimension]
reg_weights: a float32 tensor with shape [num_anchors]
match: a matcher.Match object encoding the match between anchors and
groundtruth boxes, with rows corresponding to groundtruth boxes
and columns corresponding to anchors.
Raises:
ValueError: if anchors or groundtruth_boxes are not of type
box_list.BoxList
"""
if not isinstance(anchors, box_list.BoxList):
raise ValueError('anchors must be an BoxList')
if not isinstance(groundtruth_boxes, box_list.BoxList):
raise ValueError('groundtruth_boxes must be an BoxList')
if groundtruth_labels is None:
groundtruth_labels = tf.ones(
tf.expand_dims(groundtruth_boxes.num_boxes(), 0))
groundtruth_labels = tf.expand_dims(groundtruth_labels, -1)
unmatched_shape_assert = shape_utils.assert_shape_equal(
shape_utils.combined_static_and_dynamic_shape(groundtruth_labels)
[1:],
shape_utils.combined_static_and_dynamic_shape(
self._unmatched_cls_target))
labels_and_box_shapes_assert = shape_utils.assert_shape_equal(
shape_utils.combined_static_and_dynamic_shape(groundtruth_labels)
[:1],
shape_utils.combined_static_and_dynamic_shape(
groundtruth_boxes.get())[:1])
if groundtruth_weights is None:
num_gt_boxes = groundtruth_boxes.num_boxes_static()
if not num_gt_boxes:
num_gt_boxes = groundtruth_boxes.num_boxes()
groundtruth_weights = tf.ones([num_gt_boxes], dtype=tf.float32)
with tf.control_dependencies(
[unmatched_shape_assert, labels_and_box_shapes_assert]):
match_quality_matrix = self._similarity_calc.compare(
groundtruth_boxes, anchors)
match = self._matcher.match(match_quality_matrix, **params)
reg_targets = self._create_regression_targets(
anchors, groundtruth_boxes, match)
cls_targets = self._create_classification_targets(
groundtruth_labels, match)
reg_weights = self._create_regression_weights(
match, groundtruth_weights)
cls_weights = self._create_classification_weights(
match, groundtruth_weights)
num_anchors = anchors.num_boxes_static()
if num_anchors is not None:
reg_targets = self._reset_target_shape(reg_targets, num_anchors)
cls_targets = self._reset_target_shape(cls_targets, num_anchors)
reg_weights = self._reset_target_shape(reg_weights, num_anchors)
cls_weights = self._reset_target_shape(cls_weights, num_anchors)
return cls_targets, cls_weights, reg_targets, reg_weights, match