Skip to content

GitLab

Commit 78d5f8f8 authored by Zhichao Lu's avatar Zhichao Lu Committed by lzc5123016
Browse files

Merged commit includes the following changes: 

187187978  by Zhichao Lu:

    Only updating hyperparameters if they have non-null values.

--
187097690  by Zhichao Lu:

    Rewrite some conditions a bit more clearly.

--
187085190  by Zhichao Lu:

    More informative error message.

--
186935376  by Zhichao Lu:

    Added option to evaluator.evaluate to use custom evaluator objects.

--
186808249  by Zhichao Lu:

    Fix documentation re: number of stages.

--
186775014  by Zhichao Lu:

    Change anchor generator interface to return a list of BoxLists containing anchors for different feature map layers.

--
186729028  by Zhichao Lu:

    Minor fixes to object detection.

--
186723716  by Zhichao Lu:

    Fix tf_example_decoder.py initailization issue.

--
186668505  by Zhichao Lu:

    Remove unused import.

--
186475361  by Zhichao Lu:

    Update the box predictor interface to return list of predictions - one from each feature map - instead of stacking them into one large tensor.

--
186410844  by Zhichao Lu:

    Fix PythonPath Dependencies.

--
186365384  by Zhichao Lu:

    Made some of the functions in exporter public so they can be reused.

--
186341438  by Zhichao Lu:

    Re-introducing check that label-map-path must be a valid (non-empty) string prior to overwriting pipeline config.

--
186036984  by Zhichao Lu:

    Adding default hyperparameters and allowing for overriding them via flags.

--
186026006  by Zhichao Lu:

    Strip `eval_` prefix from name argument give to TPUEstimator.evaluate since it adds the same prefix internally.

--
186016042  by Zhichao Lu:

    Add an option to evaluate models on training data.

--
185944986  by Zhichao Lu:

    let _update_label_map_path go through even if the path is empty

--
185860781  by Zhichao Lu:

    Add random normal initializer option to hyperparams builder.

    Scale the regression losses outside of the box encoder by adjusting huber loss delta and regression loss weight.

--
185846325  by Zhichao Lu:

    Add an option to normalize localization loss by the code size(number of box coordinates) in SSD Meta architecture.

--
185761217  by Zhichao Lu:

    Change multiscale_grid_anchor_generator to return anchors in normalized coordinates by default and add option to configure it.

    In SSD meta architecture, TargetAssigner operates in normalized coordinate space (i.e, groundtruth boxes are in normalized coordinates) hence we need the option to generate anchors in normalized coordinates.

--
185747733  by Zhichao Lu:

    Change the smooth L1 localization implementationt to use tf.losses.huber_loss and expose the delta parameter in the proto.

--
185715309  by Zhichao Lu:

    Obviates the need for prepadding on mobilenet v1 and v2 for fully convolutional models.

--
185685695  by Zhichao Lu:

    Fix manual stepping schedule to return first rate when there are no boundaries

--
185621650  by Zhichao Lu:

    Added target assigner proto for configuring negative class weights.

--

PiperOrigin-RevId: 187187978
parent 629adffa
Show whitespace changes
Inline Side-by-side
Showing with 235 additions and 937 deletions
research/object_detection/BUILD deleted 100644 → 0
View file @ 629adffa
# Tensorflow Object Detection API: main runnables.
package(
default_visibility = ["//visibility:public"],
)
load("//learning/brain/contrib/learn/tpu:tpu.bzl", "cloud_tpu_py_binaries")
licenses(["notice"])
# Apache 2.0
exports_files(["LICENSE"])
py_library(
name = "inputs",
srcs = [
"inputs.py",
],
deps = [
"//tensorflow",
"//tensorflow/models/research/object_detection/builders:dataset_builder",
"//tensorflow/models/research/object_detection/builders:image_resizer_builder",
"//tensorflow/models/research/object_detection/builders:model_builder",
"//tensorflow/models/research/object_detection/builders:preprocessor_builder",
"//tensorflow/models/research/object_detection/protos:input_reader_py_pb2",
"//tensorflow/models/research/object_detection/protos:model_py_pb2",
"//tensorflow/models/research/object_detection/protos:train_py_pb2",
"//tensorflow/models/research/object_detection/utils:config_util",
"//tensorflow/models/research/object_detection/utils:dataset_util",
"//tensorflow/models/research/object_detection/utils:ops",
],
)
py_test(
name = "inputs_test",
srcs = [
"inputs_test.py",
],
data = [
"//tensorflow/models/research/object_detection/data:pet_label_map.pbtxt",
"//tensorflow/models/research/object_detection/samples/configs:faster_rcnn_resnet50_pets.config",
"//tensorflow/models/research/object_detection/samples/configs:ssd_inception_v2_pets.config",
"//tensorflow/models/research/object_detection/test_data:pets_examples.record",
],
deps = [
":inputs",
"//tensorflow",
"//tensorflow/models/research/object_detection/core:standard_fields",
"//tensorflow/models/research/object_detection/utils:config_util",
],
)
py_binary(
name = "model",
srcs = [
"model.py",
],
deps = [
":inputs",
":model_hparams",
"//tensorflow",
"//tensorflow/models/research/object_detection:eval_util",
"//tensorflow/models/research/object_detection/builders:model_builder",
"//tensorflow/models/research/object_detection/builders:optimizer_builder",
"//tensorflow/models/research/object_detection/metrics:coco_evaluation",
"//tensorflow/models/research/object_detection/utils:config_util",
"//tensorflow/models/research/object_detection/utils:label_map_util",
"//tensorflow/models/research/object_detection/utils:ops",
"//tensorflow/models/research/object_detection/utils:shape_utils",
"//tensorflow/models/research/object_detection/utils:variables_helper",
"//tensorflow/models/research/object_detection/utils:visualization_utils",
],
)
py_library(
name = "model_hparams",
srcs = [
"model_hparams.py",
],
deps = [
"//tensorflow",
],
)
py_test(
name = "model_test",
timeout = "long",
srcs = [
"model_test.py",
],
data = [
"//tensorflow/models/research/object_detection/data:pet_label_map.pbtxt",
"//tensorflow/models/research/object_detection/samples/configs:faster_rcnn_resnet50_pets.config",
"//tensorflow/models/research/object_detection/samples/configs:ssd_inception_v2_pets.config",
"//tensorflow/models/research/object_detection/test_data:pets_examples.record",
],
deps = [
":inputs",
":model",
":model_hparams",
":model_test_util",
"//mock",
"//tensorflow",
"//tensorflow/models/research/object_detection/core:standard_fields",
"//tensorflow/models/research/object_detection/data_decoders:tf_example_decoder",
"//tensorflow/models/research/object_detection/utils:config_util",
"//tensorflow/models/research/object_detection/utils:ops",
],
)
MODEL_TPU_DEPS = [
":inputs",
":model",
":model_hparams",
"//tensorflow",
"//tensorflow/models/research/object_detection:eval_util",
"//tensorflow/models/research/object_detection/builders:model_builder",
"//tensorflow/models/research/object_detection/builders:optimizer_builder",
"//tensorflow/models/research/object_detection/metrics:coco_evaluation",
"//tensorflow/models/research/object_detection/utils:config_util",
"//tensorflow/models/research/object_detection/utils:label_map_util",
"//tensorflow/models/research/object_detection/utils:ops",
"//tensorflow/models/research/object_detection/utils:variables_helper",
"//tensorflow/models/research/object_detection/utils:visualization_utils",
]
cloud_tpu_py_binaries(
name = "model_tpu",
srcs = [
"model_tpu.py",
],
main = "model_tpu.py",
deps = MODEL_TPU_DEPS,
)
py_library(
name = "model_tpu_lib",
srcs = [
"model_tpu.py",
],
deps = MODEL_TPU_DEPS,
)
py_library(
name = "model_test_util",
srcs = [
"model_test_util.py",
],
deps = [
":model",
":model_hparams",
"//tensorflow",
],
)
py_binary(
name = "train",
srcs = [
"train.py",
],
deps = [
":trainer",
"//tensorflow",
"//tensorflow/models/research/object_detection/builders:dataset_builder",
"//tensorflow/models/research/object_detection/builders:model_builder",
"//tensorflow/models/research/object_detection/utils:config_util",
"//tensorflow/models/research/object_detection/utils:dataset_util",
],
)
py_library(
name = "trainer",
srcs = ["trainer.py"],
deps = [
"//tensorflow",
"//tensorflow/models/research/object_detection/builders:optimizer_builder",
"//tensorflow/models/research/object_detection/builders:preprocessor_builder",
"//tensorflow/models/research/object_detection/core:batcher",
"//tensorflow/models/research/object_detection/core:preprocessor",
"//tensorflow/models/research/object_detection/core:standard_fields",
"//tensorflow/models/research/object_detection/utils:ops",
"//tensorflow/models/research/object_detection/utils:variables_helper",
"//third_party/tensorflow_models/slim:model_deploy",
],
)
py_test(
name = "trainer_test",
srcs = ["trainer_test.py"],
deps = [
":trainer",
"//tensorflow",
"//tensorflow/models/research/object_detection/core:losses",
"//tensorflow/models/research/object_detection/core:model",
"//tensorflow/models/research/object_detection/core:standard_fields",
"//tensorflow/models/research/object_detection/protos:train_py_pb2",
],
)
py_library(
name = "eval_util",
srcs = [
"eval_util.py",
],
deps = [
"//tensorflow",
"//tensorflow/models/research/object_detection/core:box_list",
"//tensorflow/models/research/object_detection/core:box_list_ops",
"//tensorflow/models/research/object_detection/core:keypoint_ops",
"//tensorflow/models/research/object_detection/core:standard_fields",
"//tensorflow/models/research/object_detection/utils:label_map_util",
"//tensorflow/models/research/object_detection/utils:ops",
"//tensorflow/models/research/object_detection/utils:visualization_utils",
],
)
py_library(
name = "evaluator",
srcs = ["evaluator.py"],
deps = [
"//tensorflow",
"//tensorflow/models/research/object_detection:eval_util",
"//tensorflow/models/research/object_detection/core:prefetcher",
"//tensorflow/models/research/object_detection/core:standard_fields",
"//tensorflow/models/research/object_detection/metrics:coco_evaluation",
"//tensorflow/models/research/object_detection/protos:eval_py_pb2",
"//tensorflow/models/research/object_detection/utils:object_detection_evaluation",
],
)
py_binary(
name = "eval",
srcs = [
"eval.py",
],
deps = [
":evaluator",
"//tensorflow",
"//tensorflow/models/research/object_detection/builders:dataset_builder",
"//tensorflow/models/research/object_detection/builders:model_builder",
"//tensorflow/models/research/object_detection/utils:config_util",
"//tensorflow/models/research/object_detection/utils:dataset_util",
"//tensorflow/models/research/object_detection/utils:label_map_util",
],
)
py_library(
name = "exporter",
srcs = [
"exporter.py",
],
deps = [
"//tensorflow",
"//tensorflow/python/tools:freeze_graph_lib",
"//tensorflow/models/research/object_detection/builders:model_builder",
"//tensorflow/models/research/object_detection/core:standard_fields",
"//tensorflow/models/research/object_detection/data_decoders:tf_example_decoder",
],
)
py_test(
name = "exporter_test",
srcs = [
"exporter_test.py",
],
deps = [
":exporter",
"//tensorflow",
"//tensorflow/models/research/object_detection/builders:model_builder",
"//tensorflow/models/research/object_detection/core:model",
"//tensorflow/models/research/object_detection/protos:pipeline_py_pb2",
],
)
py_binary(
name = "export_inference_graph",
srcs = [
"export_inference_graph.py",
],
deps = [
":exporter",
"//tensorflow",
"//tensorflow/models/research/object_detection/protos:pipeline_py_pb2",
],
)
research/object_detection/anchor_generators/BUILD deleted 100644 → 0
View file @ 629adffa
# Tensorflow Object Detection API: Anchor Generator implementations.
package(
default_visibility = ["//visibility:public"],
)
licenses(["notice"])
# Apache 2.0
py_library(
name = "grid_anchor_generator",
srcs = [
"grid_anchor_generator.py",
],
deps = [
"//tensorflow",
"//tensorflow/models/research/object_detection/core:anchor_generator",
"//tensorflow/models/research/object_detection/core:box_list",
"//tensorflow/models/research/object_detection/utils:ops",
],
)
py_test(
name = "grid_anchor_generator_test",
srcs = [
"grid_anchor_generator_test.py",
],
deps = [
":grid_anchor_generator",
"//tensorflow",
"//tensorflow/models/research/object_detection/utils:test_case",
],
)
py_library(
name = "multiple_grid_anchor_generator",
srcs = [
"multiple_grid_anchor_generator.py",
],
deps = [
":grid_anchor_generator",
"//tensorflow",
"//tensorflow/models/research/object_detection/core:anchor_generator",
"//tensorflow/models/research/object_detection/core:box_list_ops",
],
)
py_test(
name = "multiple_grid_anchor_generator_test",
srcs = [
"multiple_grid_anchor_generator_test.py",
],
deps = [
":multiple_grid_anchor_generator",
"//numpy",
"//tensorflow/models/research/object_detection/utils:test_case",
],
)
py_library(
name = "multiscale_grid_anchor_generator",
srcs = [
"multiscale_grid_anchor_generator.py",
],
deps = [
":grid_anchor_generator",
"//tensorflow",
"//tensorflow/models/research/object_detection/core:anchor_generator",
"//tensorflow/models/research/object_detection/core:box_list_ops",
],
)
py_test(
name = "multiscale_grid_anchor_generator_test",
srcs = [
"multiscale_grid_anchor_generator_test.py",
],
deps = [
":multiscale_grid_anchor_generator",
"//numpy",
"//tensorflow/models/research/object_detection/utils:test_case",
],
)
research/object_detection/anchor_generators/grid_anchor_generator.py
View file @ 78d5f8f8
......@@ -93,11 +93,9 @@ class GridAnchorGenerator(anchor_generator.AnchorGenerator):
allowed.
Returns:
boxes: a BoxList holding a collection of N anchor boxes. Additionally
this BoxList also holds a `feature_map_index` field which is set to 0
for each anchor; this field exists for interchangeability reasons with
the MultipleGridAnchorGenerator (see the docstring for the corresponding
`_generate` function in multiple_grid_anchor_generator.py)
boxes_list: a list of BoxLists each holding anchor boxes corresponding to
the input feature map shapes.
Raises:
ValueError: if feature_map_shape_list, box_specs_list do not have the same
length.
......@@ -128,7 +126,7 @@ class GridAnchorGenerator(anchor_generator.AnchorGenerator):
num_anchors = anchors.num_boxes()
anchor_indices = tf.zeros([num_anchors])
anchors.add_field('feature_map_index', anchor_indices)
return anchors
return [anchors]
def tile_anchors(grid_height,
......
research/object_detection/anchor_generators/grid_anchor_generator_test.py
View file @ 78d5f8f8
......@@ -31,8 +31,8 @@ class GridAnchorGeneratorTest(test_case.TestCase):
anchor_offset = [7, -3]
anchor_generator = grid_anchor_generator.GridAnchorGenerator(
scales, aspect_ratios, anchor_offset=anchor_offset)
anchors = anchor_generator.generate(feature_map_shape_list=[(1, 1)])
anchor_corners = anchors.get()
anchors_list = anchor_generator.generate(feature_map_shape_list=[(1, 1)])
anchor_corners = anchors_list[0].get()
return (anchor_corners,)
exp_anchor_corners = [[-121, -35, 135, 29], [-249, -67, 263, 61],
[-505, -131, 519, 125], [-57, -67, 71, 61],
......@@ -57,8 +57,8 @@ class GridAnchorGeneratorTest(test_case.TestCase):
anchor_stride=anchor_stride,
anchor_offset=anchor_offset)
anchors = anchor_generator.generate(feature_map_shape_list=[(2, 2)])
anchor_corners = anchors.get()
anchors_list = anchor_generator.generate(feature_map_shape_list=[(2, 2)])
anchor_corners = anchors_list[0].get()
return (anchor_corners,)
exp_anchor_corners = [[-2.5, -2.5, 2.5, 2.5], [-5., -5., 5., 5.],
[-10., -10., 10., 10.], [-2.5, 16.5, 2.5, 21.5],
......@@ -83,9 +83,9 @@ class GridAnchorGeneratorTest(test_case.TestCase):
anchor_stride=anchor_stride,
anchor_offset=anchor_offset)
anchors = anchor_generator.generate(
anchors_list = anchor_generator.generate(
feature_map_shape_list=[(feature_map_height, feature_map_width)])
anchor_corners = anchors.get()
anchor_corners = anchors_list[0].get()
return (anchor_corners,)
exp_anchor_corners = [[-2.5, -2.5, 2.5, 2.5], [-5., -5., 5., 5.],
......
research/object_detection/anchor_generators/multiple_grid_anchor_generator.py
View file @ 78d5f8f8
......@@ -165,10 +165,9 @@ class MultipleGridAnchorGenerator(anchor_generator.AnchorGenerator):
grid.
Returns:
boxes: a BoxList holding a collection of N anchor boxes. Additionally
this BoxList also holds a `feature_map_index` field which, for each
anchor, stores the index of the corresponding feature map which was used
to generate it.
boxes_list: a list of BoxLists each holding anchor boxes corresponding to
the input feature map shapes.
Raises:
ValueError: if feature_map_shape_list, box_specs_list do not have the same
length.
......@@ -211,7 +210,6 @@ class MultipleGridAnchorGenerator(anchor_generator.AnchorGenerator):
raise ValueError('%s must be a list of pairs.' % arg_name)
anchor_grid_list = []
anchor_indices_list = []
min_im_shape = tf.minimum(im_height, im_width)
scale_height = min_im_shape / im_height
scale_width = min_im_shape / im_width
......@@ -219,10 +217,11 @@ class MultipleGridAnchorGenerator(anchor_generator.AnchorGenerator):
scale_height * self._base_anchor_size[0],
scale_width * self._base_anchor_size[1]
]
for feature_map_index, (
grid_size, scales, aspect_ratios, stride, offset) in enumerate(
zip(feature_map_shape_list, self._scales, self._aspect_ratios,
anchor_strides, anchor_offsets)):
for feature_map_index, (grid_size, scales, aspect_ratios, stride,
offset) in enumerate(
zip(feature_map_shape_list, self._scales,
self._aspect_ratios, anchor_strides,
anchor_offsets)):
tiled_anchors = grid_anchor_generator.tile_anchors(
grid_height=grid_size[0],
grid_width=grid_size[1],
......@@ -231,30 +230,17 @@ class MultipleGridAnchorGenerator(anchor_generator.AnchorGenerator):
base_anchor_size=base_anchor_size,
anchor_stride=stride,
anchor_offset=offset)
anchor_grid_list.append(tiled_anchors)
if self._clip_window is not None:
tiled_anchors = box_list_ops.clip_to_window(
tiled_anchors, self._clip_window, filter_nonoverlapping=False)
num_anchors_in_layer = tiled_anchors.num_boxes_static()
if num_anchors_in_layer is None:
num_anchors_in_layer = tiled_anchors.num_boxes()
anchor_indices_list.append(
feature_map_index * tf.ones([num_anchors_in_layer]))
concatenated_anchors = box_list_ops.concatenate(anchor_grid_list)
anchor_indices = tf.concat(anchor_indices_list, 0)
num_anchors = concatenated_anchors.num_boxes_static()
if num_anchors is None:
num_anchors = concatenated_anchors.num_boxes()
if self._clip_window is not None:
concatenated_anchors = box_list_ops.clip_to_window(
concatenated_anchors, self._clip_window, filter_nonoverlapping=False)
# TODO: make reshape an option for the clip_to_window op
concatenated_anchors.set(
tf.reshape(concatenated_anchors.get(), [num_anchors, 4]))
stddevs_tensor = 0.01 * tf.ones(
[num_anchors, 4], dtype=tf.float32, name='stddevs')
concatenated_anchors.add_field('stddev', stddevs_tensor)
concatenated_anchors.add_field('feature_map_index', anchor_indices)
anchor_indices = feature_map_index * tf.ones([num_anchors_in_layer])
tiled_anchors.add_field('feature_map_index', anchor_indices)
anchor_grid_list.append(tiled_anchors)
return concatenated_anchors
return anchor_grid_list
def create_ssd_anchors(num_layers=6,
......@@ -285,7 +271,7 @@ def create_ssd_anchors(num_layers=6,
grid sizes passed in at generation time)
min_scale: scale of anchors corresponding to finest resolution (float)
max_scale: scale of anchors corresponding to coarsest resolution (float)
scales: As list of anchor scales to use. When not None and not emtpy,
scales: As list of anchor scales to use. When not None and not empty,
min_scale and max_scale are not used.
aspect_ratios: list or tuple of (float) aspect ratios to place on each
grid point.
......
research/object_detection/anchor_generators/multiple_grid_anchor_generator_test.py
View file @ 78d5f8f8
......@@ -37,8 +37,8 @@ class MultipleGridAnchorGeneratorTest(test_case.TestCase):
base_anchor_size=tf.constant([256, 256], dtype=tf.float32),
anchor_strides=[(16, 16)],
anchor_offsets=[(7, -3)])
anchors = anchor_generator.generate(feature_map_shape_list=[(1, 1)])
return anchors.get()
anchors_list = anchor_generator.generate(feature_map_shape_list=[(1, 1)])
return anchors_list[0].get()
exp_anchor_corners = [[-121, -35, 135, 29], [-249, -67, 263, 61],
[-505, -131, 519, 125], [-57, -67, 71, 61],
[-121, -131, 135, 125], [-249, -259, 263, 253],
......@@ -57,8 +57,8 @@ class MultipleGridAnchorGeneratorTest(test_case.TestCase):
base_anchor_size=tf.constant([10, 10], dtype=tf.float32),
anchor_strides=[(19, 19)],
anchor_offsets=[(0, 0)])
anchors = anchor_generator.generate(feature_map_shape_list=[(2, 2)])
return anchors.get()
anchors_list = anchor_generator.generate(feature_map_shape_list=[(2, 2)])
return anchors_list[0].get()
exp_anchor_corners = [[-2.5, -2.5, 2.5, 2.5], [-5., -5., 5., 5.],
[-10., -10., 10., 10.], [-2.5, 16.5, 2.5, 21.5],
[-5., 14., 5, 24], [-10., 9., 10, 29],
......@@ -76,9 +76,9 @@ class MultipleGridAnchorGeneratorTest(test_case.TestCase):
anchor_generator = ag.MultipleGridAnchorGenerator(
box_specs_list, base_anchor_size=tf.constant([1, 1],
dtype=tf.float32))
anchors = anchor_generator.generate(feature_map_shape_list=[(tf.constant(
1, dtype=tf.int32), tf.constant(2, dtype=tf.int32))])
return anchors.get()
anchors_list = anchor_generator.generate(feature_map_shape_list=[(
tf.constant(1, dtype=tf.int32), tf.constant(2, dtype=tf.int32))])
return anchors_list[0].get()
exp_anchor_corners = [[0., -0.25, 1., 0.75], [0., 0.25, 1., 1.25]]
anchor_corners_out = self.execute(graph_fn, [])
......@@ -91,9 +91,9 @@ class MultipleGridAnchorGeneratorTest(test_case.TestCase):
anchor_generator = ag.MultipleGridAnchorGenerator(
box_specs_list, base_anchor_size=tf.constant([1, 1],
dtype=tf.float32))
anchors = anchor_generator.generate(feature_map_shape_list=[(height,
anchors_list = anchor_generator.generate(feature_map_shape_list=[(height,
width)])
return anchors.get()
return anchors_list[0].get()
exp_anchor_corners = [[0., -0.25, 1., 0.75], [0., 0.25, 1., 1.25]]
......@@ -109,12 +109,12 @@ class MultipleGridAnchorGeneratorTest(test_case.TestCase):
anchor_generator = ag.MultipleGridAnchorGenerator(
box_specs_list, base_anchor_size=tf.constant([1, 1],
dtype=tf.float32))
anchors = anchor_generator.generate(
anchors_list = anchor_generator.generate(
feature_map_shape_list=[(tf.constant(1, dtype=tf.int32), tf.constant(
2, dtype=tf.int32))],
im_height=320,
im_width=640)
return anchors.get()
return anchors_list[0].get()
exp_anchor_corners = [[0., 0., 1., 0.5], [0., 0.5, 1., 1.]]
anchor_corners_out = self.execute(graph_fn, [])
......@@ -131,9 +131,9 @@ class MultipleGridAnchorGeneratorTest(test_case.TestCase):
base_anchor_size=tf.constant([1.0, 1.0], dtype=tf.float32),
anchor_strides=[(.25, .25), (.5, .5)],
anchor_offsets=[(.125, .125), (.25, .25)])
anchors = anchor_generator.generate(feature_map_shape_list=[(4, 4),
(2, 2)])
return anchors.get()
anchors_list = anchor_generator.generate(feature_map_shape_list=[(4, 4), (
2, 2)])
return [anchors.get() for anchors in anchors_list]
# height and width of box with .5 aspect ratio
h = np.sqrt(2)
w = 1.0/np.sqrt(2)
......@@ -150,7 +150,7 @@ class MultipleGridAnchorGeneratorTest(test_case.TestCase):
[.125-1.0, .125-1.0, .125+1.0, .125+1.0],
[.125-.5*h, .125-.5*w, .125+.5*h, .125+.5*w],]
anchor_corners_out = self.execute(graph_fn, [])
anchor_corners_out = np.concatenate(self.execute(graph_fn, []), axis=0)
self.assertEquals(anchor_corners_out.shape, (56, 4))
big_grid_corners = anchor_corners_out[0:3, :]
small_grid_corners = anchor_corners_out[48:, :]
......@@ -168,9 +168,9 @@ class MultipleGridAnchorGeneratorTest(test_case.TestCase):
box_specs_list,
base_anchor_size=tf.constant([1.0, 1.0], dtype=tf.float32),
clip_window=clip_window)
anchors = anchor_generator.generate(feature_map_shape_list=[(4, 4),
(2, 2)])
return anchors.get()
anchors_list = anchor_generator.generate(feature_map_shape_list=[(4, 4), (
2, 2)])
return [anchors.get() for anchors in anchors_list]
# height and width of box with .5 aspect ratio
h = np.sqrt(2)
w = 1.0/np.sqrt(2)
......@@ -183,7 +183,7 @@ class MultipleGridAnchorGeneratorTest(test_case.TestCase):
[.25, .25, 1, 1],
[.75-.5*h, .75-.5*w, 1, 1]]
anchor_corners_out = self.execute(graph_fn, [])
anchor_corners_out = np.concatenate(self.execute(graph_fn, []), axis=0)
small_grid_corners = anchor_corners_out[48:, :]
self.assertAllClose(small_grid_corners, exp_small_grid_corners)
......@@ -264,10 +264,10 @@ class CreateSSDAnchorsTest(test_case.TestCase):
feature_map_shape_list = [(38, 38), (19, 19), (10, 10),
(5, 5), (3, 3), (1, 1)]
anchors = anchor_generator.generate(
anchors_list = anchor_generator.generate(
feature_map_shape_list=feature_map_shape_list)
return anchors.get()
anchor_corners_out = self.execute(graph_fn1, [])
return [anchors.get() for anchors in anchors_list]
anchor_corners_out = np.concatenate(self.execute(graph_fn1, []), axis=0)
self.assertEquals(anchor_corners_out.shape, (7308, 4))
def graph_fn2():
......@@ -278,10 +278,10 @@ class CreateSSDAnchorsTest(test_case.TestCase):
feature_map_shape_list = [(38, 38), (19, 19), (10, 10),
(5, 5), (3, 3), (1, 1)]
anchors = anchor_generator.generate(
anchors_list = anchor_generator.generate(
feature_map_shape_list=feature_map_shape_list)
return anchors.get()
anchor_corners_out = self.execute(graph_fn2, [])
return [anchors.get() for anchors in anchors_list]
anchor_corners_out = np.concatenate(self.execute(graph_fn2, []), axis=0)
self.assertEquals(anchor_corners_out.shape, (11640, 4))
......
research/object_detection/anchor_generators/multiscale_grid_anchor_generator.py
View file @ 78d5f8f8
......@@ -21,14 +21,15 @@ T.-Y. Lin, P. Goyal, R. Girshick, K. He, P. Dollar
"""
from object_detection.anchor_generators import grid_anchor_generator
from object_detection.core import anchor_generator
from object_detection.core import box_list_ops
class MultiscaleGridAnchorGenerator(object):
class MultiscaleGridAnchorGenerator(anchor_generator.AnchorGenerator):
"""Generate a grid of anchors for multiple CNN layers of different scale."""
def __init__(self, min_level, max_level, anchor_scale, aspect_ratios,
scales_per_octave):
scales_per_octave, normalize_coordinates=True):
"""Constructs a MultiscaleGridAnchorGenerator.
To construct anchors, at multiple scale resolutions, one must provide a
......@@ -48,10 +49,13 @@ class MultiscaleGridAnchorGenerator(object):
aspect_ratios: list or tuple of (float) aspect ratios to place on each
grid point.
scales_per_octave: integer number of intermediate scales per scale octave.
normalize_coordinates: whether to produce anchors in normalized
coordinates. (defaults to True).
"""
self._anchor_grid_info = []
self._aspect_ratios = aspect_ratios
self._scales_per_octave = scales_per_octave
self._normalize_coordinates = normalize_coordinates
for level in range(min_level, max_level + 1):
anchor_stride = [2**level, 2**level]
......@@ -80,7 +84,7 @@ class MultiscaleGridAnchorGenerator(object):
return len(self._anchor_grid_info) * [
len(self._aspect_ratios) * self._scales_per_octave]
def generate(self, feature_map_shape_list, im_height, im_width):
def _generate(self, feature_map_shape_list, im_height, im_width):
"""Generates a collection of bounding boxes to be used as anchors.
Currently we require the input image shape to be statically defined. That
......@@ -95,7 +99,8 @@ class MultiscaleGridAnchorGenerator(object):
im_width: the width of the image to generate the grid for.
Returns:
boxes: a BoxList holding a collection of N anchor boxes
boxes_list: a list of BoxLists each holding anchor boxes corresponding to
the input feature map shapes.
Raises:
ValueError: if im_height and im_width are not integers.
"""
......@@ -105,7 +110,7 @@ class MultiscaleGridAnchorGenerator(object):
anchor_grid_list = []
for feat_shape, grid_info in zip(feature_map_shape_list,
self._anchor_grid_info):
# TODO check the feature_map_shape_list is consistent with
# TODO(rathodv) check the feature_map_shape_list is consistent with
# self._anchor_grid_info
level = grid_info['level']
stride = 2**level
......@@ -123,9 +128,11 @@ class MultiscaleGridAnchorGenerator(object):
base_anchor_size=base_anchor_size,
anchor_stride=anchor_stride,
anchor_offset=anchor_offset)
anchor_grid_list.append(
ag.generate(feature_map_shape_list=[(feat_h, feat_w)]))
(anchor_grid,) = ag.generate(feature_map_shape_list=[(feat_h, feat_w)])
concatenated_anchors = box_list_ops.concatenate(anchor_grid_list)
if self._normalize_coordinates:
anchor_grid = box_list_ops.to_normalized_coordinates(
anchor_grid, im_height, im_width, check_range=False)
anchor_grid_list.append(anchor_grid)
return concatenated_anchors
return anchor_grid_list
research/object_detection/anchor_generators/multiscale_grid_anchor_generator_test.py
View file @ 78d5f8f8
......@@ -37,10 +37,35 @@ class MultiscaleGridAnchorGeneratorTest(test_case.TestCase):
[-16, -48, 112, 80],
[-16, -16, 112, 112]]
anchor_generator = mg.MultiscaleGridAnchorGenerator(
min_level, max_level, anchor_scale, aspect_ratios, scales_per_octave)
anchors = anchor_generator.generate(feature_map_shape_list,
im_height, im_width)
anchor_corners = anchors.get()
min_level, max_level, anchor_scale, aspect_ratios, scales_per_octave,
normalize_coordinates=False)
anchors_list = anchor_generator.generate(
feature_map_shape_list, im_height=im_height, im_width=im_width)
anchor_corners = anchors_list[0].get()
with self.test_session():
anchor_corners_out = anchor_corners.eval()
self.assertAllClose(anchor_corners_out, exp_anchor_corners)
def test_construct_single_anchor_in_normalized_coordinates(self):
min_level = 5
max_level = 5
anchor_scale = 4.0
aspect_ratios = [1.0]
scales_per_octave = 1
im_height = 64
im_width = 128
feature_map_shape_list = [(2, 2)]
exp_anchor_corners = [[-48./64, -48./128, 80./64, 80./128],
[-48./64, -16./128, 80./64, 112./128],
[-16./64, -48./128, 112./64, 80./128],
[-16./64, -16./128, 112./64, 112./128]]
anchor_generator = mg.MultiscaleGridAnchorGenerator(
min_level, max_level, anchor_scale, aspect_ratios, scales_per_octave,
normalize_coordinates=True)
anchors_list = anchor_generator.generate(
feature_map_shape_list, im_height=im_height, im_width=im_width)
anchor_corners = anchors_list[0].get()
with self.test_session():
anchor_corners_out = anchor_corners.eval()
......@@ -53,7 +78,8 @@ class MultiscaleGridAnchorGeneratorTest(test_case.TestCase):
aspect_ratios = [1.0, 2.0]
scales_per_octave = 3
anchor_generator = mg.MultiscaleGridAnchorGenerator(
min_level, max_level, anchor_scale, aspect_ratios, scales_per_octave)
min_level, max_level, anchor_scale, aspect_ratios, scales_per_octave,
normalize_coordinates=False)
self.assertEqual(anchor_generator.num_anchors_per_location(), [6, 6])
def test_construct_single_anchor_fails_with_tensor_image_size(self):
......@@ -66,9 +92,11 @@ class MultiscaleGridAnchorGeneratorTest(test_case.TestCase):
im_width = tf.constant(64)
feature_map_shape_list = [(2, 2)]
anchor_generator = mg.MultiscaleGridAnchorGenerator(
min_level, max_level, anchor_scale, aspect_ratios, scales_per_octave)
min_level, max_level, anchor_scale, aspect_ratios, scales_per_octave,
normalize_coordinates=False)
with self.assertRaises(ValueError):
anchor_generator.generate(feature_map_shape_list, im_height, im_width)
anchor_generator.generate(
feature_map_shape_list, im_height=im_height, im_width=im_width)
def test_construct_single_anchor_with_odd_input_dimension(self):
......@@ -82,10 +110,11 @@ class MultiscaleGridAnchorGeneratorTest(test_case.TestCase):
im_width = 65
feature_map_shape_list = [(3, 3)]
anchor_generator = mg.MultiscaleGridAnchorGenerator(
min_level, max_level, anchor_scale, aspect_ratios, scales_per_octave)
anchors = anchor_generator.generate(feature_map_shape_list, im_height,
im_width)
anchor_corners = anchors.get()
min_level, max_level, anchor_scale, aspect_ratios, scales_per_octave,
normalize_coordinates=False)
anchors_list = anchor_generator.generate(
feature_map_shape_list, im_height=im_height, im_width=im_width)
anchor_corners = anchors_list[0].get()
return (anchor_corners,)
anchor_corners_out = self.execute(graph_fn, [])
exp_anchor_corners = [[-64, -64, 64, 64],
......@@ -111,13 +140,15 @@ class MultiscaleGridAnchorGeneratorTest(test_case.TestCase):
im_width = 64
feature_map_shape_list = [(2, 2), (1, 1)]
anchor_generator = mg.MultiscaleGridAnchorGenerator(
min_level, max_level, anchor_scale, aspect_ratios, scales_per_octave)
anchors = anchor_generator.generate(feature_map_shape_list, im_height,
im_width)
anchor_corners = anchors.get()
return (anchor_corners,)
min_level, max_level, anchor_scale, aspect_ratios, scales_per_octave,
normalize_coordinates=False)
anchors_list = anchor_generator.generate(feature_map_shape_list,
im_height=im_height,
im_width=im_width)
anchor_corners = [anchors.get() for anchors in anchors_list]
return anchor_corners
anchor_corners_out = self.execute(graph_fn, [])
anchor_corners_out = np.concatenate(self.execute(graph_fn, []), axis=0)
exp_anchor_corners = [[-48, -48, 80, 80],
[-48, -16, 80, 112],
[-16, -48, 112, 80],
......@@ -135,19 +166,22 @@ class MultiscaleGridAnchorGeneratorTest(test_case.TestCase):
scales_per_octave = 2
im_height = 64
im_width = 64
feature_map_shape_list = [(1, 1), (1, 1)]
feature_map_shape_list = [(1, 1)]
anchor_generator = mg.MultiscaleGridAnchorGenerator(
min_level, max_level, anchor_scale, aspect_ratios, scales_per_octave)
anchors = anchor_generator.generate(feature_map_shape_list, im_height,
im_width)
anchor_corners = anchors.get()
return (anchor_corners,)
min_level, max_level, anchor_scale, aspect_ratios, scales_per_octave,
normalize_coordinates=False)
anchors_list = anchor_generator.generate(feature_map_shape_list,
im_height=im_height,
im_width=im_width)
anchor_corners = [anchors.get() for anchors in anchors_list]
return anchor_corners
# There are 4 set of anchors in this configuration. The order is:
# [[2**0.0 intermediate scale + 1.0 aspect],
# [2**0.5 intermediate scale + 1.0 aspect]]
exp_anchor_corners = [[-96., -96., 160., 160.],
[-149.0193, -149.0193, 213.0193, 213.0193]]
anchor_corners_out = self.execute(graph_fn, [])
self.assertAllClose(anchor_corners_out, exp_anchor_corners)
......@@ -160,18 +194,21 @@ class MultiscaleGridAnchorGeneratorTest(test_case.TestCase):
scales_per_octave = 2
im_height = 64
im_width = 64
feature_map_shape_list = [(1, 1), (1, 1), (1, 1), (1, 1)]
feature_map_shape_list = [(1, 1)]
anchor_generator = mg.MultiscaleGridAnchorGenerator(
min_level, max_level, anchor_scale, aspect_ratios, scales_per_octave)
anchors = anchor_generator.generate(feature_map_shape_list, im_height,
im_width)
anchor_corners = anchors.get()
min_level, max_level, anchor_scale, aspect_ratios, scales_per_octave,
normalize_coordinates=False)
anchors_list = anchor_generator.generate(feature_map_shape_list,
im_height=im_height,
im_width=im_width)
anchor_corners = [anchors.get() for anchors in anchors_list]
return anchor_corners
# There are 4 set of anchors in this configuration. The order is:
# [[2**0.0 intermediate scale + 1.0 aspect],
# [2**0.5 intermediate scale + 1.0 aspect],
# [2**0.0 intermediate scale + 2.0 aspect],
# [2**0.5 intermediate scale + 2.0 aspect]]
exp_anchor_corners = [[-96., -96., 160., 160.],
[-149.0193, -149.0193, 213.0193, 213.0193],
[-58.50967, -149.0193, 122.50967, 213.0193],
......@@ -193,18 +230,22 @@ class MultiscaleGridAnchorGeneratorTest(test_case.TestCase):
feature_map_shape_list = [(feature_map1_height, feature_map1_width),
(feature_map2_height, feature_map2_width)]
anchor_generator = mg.MultiscaleGridAnchorGenerator(
min_level, max_level, anchor_scale, aspect_ratios, scales_per_octave)
anchors = anchor_generator.generate(feature_map_shape_list, im_height,
im_width)
anchor_corners = anchors.get()
return (anchor_corners,)
min_level, max_level, anchor_scale, aspect_ratios, scales_per_octave,
normalize_coordinates=False)
anchors_list = anchor_generator.generate(feature_map_shape_list,
im_height=im_height,
im_width=im_width)
anchor_corners = [anchors.get() for anchors in anchors_list]
return anchor_corners
anchor_corners_out = self.execute_cpu(graph_fn, [
anchor_corners_out = np.concatenate(
self.execute_cpu(graph_fn, [
np.array(2, dtype=np.int32),
np.array(2, dtype=np.int32),
np.array(1, dtype=np.int32),
np.array(1, dtype=np.int32)
])
]),
axis=0)
exp_anchor_corners = [[-48, -48, 80, 80],
[-48, -16, 80, 112],
[-16, -48, 112, 80],
......
research/object_detection/box_coders/BUILD deleted 100644 → 0
View file @ 629adffa
# Tensorflow Object Detection API: Box Coder implementations.
package(
default_visibility = ["//visibility:public"],
)
licenses(["notice"])
# Apache 2.0
py_library(
name = "faster_rcnn_box_coder",
srcs = [
"faster_rcnn_box_coder.py",
],
deps = [
"//tensorflow/models/research/object_detection/core:box_coder",
"//tensorflow/models/research/object_detection/core:box_list",
],
)
py_test(
name = "faster_rcnn_box_coder_test",
srcs = [
"faster_rcnn_box_coder_test.py",
],
deps = [
":faster_rcnn_box_coder",
"//tensorflow",
"//tensorflow/models/research/object_detection/core:box_list",
],
)
py_library(
name = "keypoint_box_coder",
srcs = [
"keypoint_box_coder.py",
],
deps = [
"//tensorflow/models/research/object_detection/core:box_coder",
"//tensorflow/models/research/object_detection/core:box_list",
"//tensorflow/models/research/object_detection/core:standard_fields",
],
)
py_test(
name = "keypoint_box_coder_test",
srcs = [
"keypoint_box_coder_test.py",
],
deps = [
":keypoint_box_coder",
"//tensorflow",
"//tensorflow/models/research/object_detection/core:box_list",
"//tensorflow/models/research/object_detection/core:standard_fields",
],
)
py_library(
name = "mean_stddev_box_coder",
srcs = [
"mean_stddev_box_coder.py",
],
deps = [
"//tensorflow/models/research/object_detection/core:box_coder",
"//tensorflow/models/research/object_detection/core:box_list",
],
)
py_test(
name = "mean_stddev_box_coder_test",
srcs = [
"mean_stddev_box_coder_test.py",
],
deps = [
":mean_stddev_box_coder",
"//tensorflow",
"//tensorflow/models/research/object_detection/core:box_list",
],
)
py_library(
name = "square_box_coder",
srcs = [
"square_box_coder.py",
],
deps = [
"//tensorflow/models/research/object_detection/core:box_coder",
"//tensorflow/models/research/object_detection/core:box_list",
],
)
py_test(
name = "square_box_coder_test",
srcs = [
"square_box_coder_test.py",
],
deps = [
":square_box_coder",
"//tensorflow",
"//tensorflow/models/research/object_detection/core:box_list",
],
)
research/object_detection/builders/BUILD deleted 100644 → 0
View file @ 629adffa
# Tensorflow Object Detection API: component builders.
package(
default_visibility = ["//visibility:public"],
)
licenses(["notice"])
# Apache 2.0
py_library(
name = "model_builder",
srcs = ["model_builder.py"],
deps = [
":anchor_generator_builder",
":box_coder_builder",
":box_predictor_builder",
":hyperparams_builder",
":image_resizer_builder",
":losses_builder",
":matcher_builder",
":post_processing_builder",
":region_similarity_calculator_builder",
"//tensorflow/models/research/object_detection/core:box_predictor",
"//tensorflow/models/research/object_detection/meta_architectures:faster_rcnn_meta_arch",
"//tensorflow/models/research/object_detection/meta_architectures:rfcn_meta_arch",
"//tensorflow/models/research/object_detection/meta_architectures:ssd_meta_arch",
"//tensorflow/models/research/object_detection/models:embedded_ssd_mobilenet_v1_feature_extractor",
"//tensorflow/models/research/object_detection/models:faster_rcnn_inception_resnet_v2_feature_extractor",
"//tensorflow/models/research/object_detection/models:faster_rcnn_inception_v2_feature_extractor",
"//tensorflow/models/research/object_detection/models:faster_rcnn_nas_feature_extractor",
"//tensorflow/models/research/object_detection/models:faster_rcnn_resnet_v1_feature_extractor",
"//tensorflow/models/research/object_detection/models:ssd_inception_v2_feature_extractor",
"//tensorflow/models/research/object_detection/models:ssd_inception_v3_feature_extractor",
"//tensorflow/models/research/object_detection/models:ssd_mobilenet_v1_feature_extractor",
"//tensorflow/models/research/object_detection/models:ssd_resnet_v1_fpn_feature_extractor",
"//tensorflow/models/research/object_detection/protos:model_py_pb2",
],
)
py_test(
name = "model_builder_test",
srcs = ["model_builder_test.py"],
deps = [
":model_builder",
"//tensorflow",
"//tensorflow/models/research/object_detection/meta_architectures:faster_rcnn_meta_arch",
"//tensorflow/models/research/object_detection/meta_architectures:ssd_meta_arch",
"//tensorflow/models/research/object_detection/models:embedded_ssd_mobilenet_v1_feature_extractor",
"//tensorflow/models/research/object_detection/models:faster_rcnn_inception_resnet_v2_feature_extractor",
"//tensorflow/models/research/object_detection/models:faster_rcnn_inception_v2_feature_extractor",
"//tensorflow/models/research/object_detection/models:faster_rcnn_nas_feature_extractor",
"//tensorflow/models/research/object_detection/models:faster_rcnn_resnet_v1_feature_extractor",
"//tensorflow/models/research/object_detection/models:ssd_inception_v2_feature_extractor",
"//tensorflow/models/research/object_detection/models:ssd_inception_v3_feature_extractor",
"//tensorflow/models/research/object_detection/models:ssd_mobilenet_v1_feature_extractor",
"//tensorflow/models/research/object_detection/models:ssd_resnet_v1_fpn_feature_extractor",
"//tensorflow/models/research/object_detection/protos:model_py_pb2",
],
)
py_library(
name = "matcher_builder",
srcs = ["matcher_builder.py"],
deps = [
"//tensorflow/models/research/object_detection/matchers:argmax_matcher",
"//tensorflow/models/research/object_detection/matchers:bipartite_matcher",
"//tensorflow/models/research/object_detection/protos:matcher_py_pb2",
],
)
py_test(
name = "matcher_builder_test",
srcs = ["matcher_builder_test.py"],
deps = [
":matcher_builder",
"//tensorflow/models/research/object_detection/matchers:argmax_matcher",
"//tensorflow/models/research/object_detection/matchers:bipartite_matcher",
"//tensorflow/models/research/object_detection/protos:matcher_py_pb2",
],
)
py_library(
name = "box_coder_builder",
srcs = ["box_coder_builder.py"],
deps = [
"//tensorflow/models/research/object_detection/box_coders:faster_rcnn_box_coder",
"//tensorflow/models/research/object_detection/box_coders:keypoint_box_coder",
"//tensorflow/models/research/object_detection/box_coders:mean_stddev_box_coder",
"//tensorflow/models/research/object_detection/box_coders:square_box_coder",
"//tensorflow/models/research/object_detection/protos:box_coder_py_pb2",
],
)
py_test(
name = "box_coder_builder_test",
srcs = ["box_coder_builder_test.py"],
deps = [
":box_coder_builder",
"//tensorflow",
"//tensorflow/models/research/object_detection/box_coders:faster_rcnn_box_coder",
"//tensorflow/models/research/object_detection/box_coders:keypoint_box_coder",
"//tensorflow/models/research/object_detection/box_coders:mean_stddev_box_coder",
"//tensorflow/models/research/object_detection/box_coders:square_box_coder",
"//tensorflow/models/research/object_detection/protos:box_coder_py_pb2",
],
)
py_library(
name = "anchor_generator_builder",
srcs = ["anchor_generator_builder.py"],
deps = [
"//tensorflow/models/research/object_detection/anchor_generators:grid_anchor_generator",
"//tensorflow/models/research/object_detection/anchor_generators:multiple_grid_anchor_generator",
"//tensorflow/models/research/object_detection/anchor_generators:multiscale_grid_anchor_generator",
"//tensorflow/models/research/object_detection/protos:anchor_generator_py_pb2",
],
)
py_test(
name = "anchor_generator_builder_test",
srcs = ["anchor_generator_builder_test.py"],
deps = [
":anchor_generator_builder",
"//tensorflow",
"//tensorflow/models/research/object_detection/anchor_generators:grid_anchor_generator",
"//tensorflow/models/research/object_detection/anchor_generators:multiple_grid_anchor_generator",
"//tensorflow/models/research/object_detection/anchor_generators:multiscale_grid_anchor_generator",
"//tensorflow/models/research/object_detection/protos:anchor_generator_py_pb2",
],
)
py_library(
name = "dataset_builder",
srcs = ["dataset_builder.py"],
deps = [
"//tensorflow",
"//tensorflow/models/research/object_detection/data_decoders:tf_example_decoder",
"//tensorflow/models/research/object_detection/protos:input_reader_py_pb2",
"//tensorflow/models/research/object_detection/utils:dataset_util",
],
)
py_test(
name = "dataset_builder_test",
srcs = [
"dataset_builder_test.py",
],
deps = [
":dataset_builder",
"//tensorflow",
"//tensorflow/models/research/object_detection/core:standard_fields",
"//tensorflow/models/research/object_detection/protos:input_reader_py_pb2",
"//tensorflow/models/research/object_detection/utils:dataset_util",
],
)
py_library(
name = "input_reader_builder",
srcs = ["input_reader_builder.py"],
deps = [
"//tensorflow",
"//tensorflow/models/research/object_detection/data_decoders:tf_example_decoder",
"//tensorflow/models/research/object_detection/protos:input_reader_py_pb2",
],
)
py_test(
name = "input_reader_builder_test",
srcs = [
"input_reader_builder_test.py",
],
deps = [
":input_reader_builder",
"//tensorflow",
"//tensorflow/models/research/object_detection/core:standard_fields",
"//tensorflow/models/research/object_detection/protos:input_reader_py_pb2",
],
)
py_library(
name = "losses_builder",
srcs = ["losses_builder.py"],
deps = [
"//tensorflow/models/research/object_detection/core:losses",
"//tensorflow/models/research/object_detection/protos:losses_py_pb2",
],
)
py_test(
name = "losses_builder_test",
srcs = ["losses_builder_test.py"],
deps = [
":losses_builder",
"//tensorflow/models/research/object_detection/core:losses",
"//tensorflow/models/research/object_detection/protos:losses_py_pb2",
],
)
py_library(
name = "optimizer_builder",
srcs = ["optimizer_builder.py"],
deps = [
"//tensorflow",
"//tensorflow/models/research/object_detection/utils:learning_schedules",
],
)
py_test(
name = "optimizer_builder_test",
srcs = ["optimizer_builder_test.py"],
deps = [
":optimizer_builder",
"//tensorflow",
"//tensorflow/models/research/object_detection/protos:optimizer_py_pb2",
],
)
py_library(
name = "post_processing_builder",
srcs = ["post_processing_builder.py"],
deps = [
"//tensorflow",
"//tensorflow/models/research/object_detection/core:post_processing",
"//tensorflow/models/research/object_detection/protos:post_processing_py_pb2",
],
)
py_test(
name = "post_processing_builder_test",
srcs = ["post_processing_builder_test.py"],
deps = [
":post_processing_builder",
"//tensorflow",
"//tensorflow/models/research/object_detection/protos:post_processing_py_pb2",
],
)
py_library(
name = "hyperparams_builder",
srcs = ["hyperparams_builder.py"],
deps = [
"//tensorflow/models/research/object_detection/protos:hyperparams_py_pb2",
],
)
py_test(
name = "hyperparams_builder_test",
srcs = ["hyperparams_builder_test.py"],
deps = [
":hyperparams_builder",
"//tensorflow",
"//tensorflow/models/research/object_detection/protos:hyperparams_py_pb2",
],
)
py_library(
name = "box_predictor_builder",
srcs = ["box_predictor_builder.py"],
deps = [
":hyperparams_builder",
"//tensorflow/models/research/object_detection/core:box_predictor",
"//tensorflow/models/research/object_detection/protos:box_predictor_py_pb2",
],
)
py_test(
name = "box_predictor_builder_test",
srcs = ["box_predictor_builder_test.py"],
deps = [
":box_predictor_builder",
":hyperparams_builder",
"//tensorflow",
"//tensorflow/models/research/object_detection/protos:box_predictor_py_pb2",
"//tensorflow/models/research/object_detection/protos:hyperparams_py_pb2",
],
)
py_library(
name = "region_similarity_calculator_builder",
srcs = ["region_similarity_calculator_builder.py"],
deps = [
"//tensorflow/models/research/object_detection/core:region_similarity_calculator",
"//tensorflow/models/research/object_detection/protos:region_similarity_calculator_py_pb2",
],
)
py_test(
name = "region_similarity_calculator_builder_test",
srcs = ["region_similarity_calculator_builder_test.py"],
deps = [
":region_similarity_calculator_builder",
"//tensorflow",
],
)
py_library(
name = "preprocessor_builder",
srcs = ["preprocessor_builder.py"],
deps = [
"//tensorflow",
"//tensorflow/models/research/object_detection/core:preprocessor",
"//tensorflow/models/research/object_detection/protos:preprocessor_py_pb2",
],
)
py_test(
name = "preprocessor_builder_test",
srcs = [
"preprocessor_builder_test.py",
],
deps = [
":preprocessor_builder",
"//tensorflow",
"//tensorflow/models/research/object_detection/core:preprocessor",
"//tensorflow/models/research/object_detection/protos:preprocessor_py_pb2",
],
)
py_library(
name = "image_resizer_builder",
srcs = ["image_resizer_builder.py"],
deps = [
"//tensorflow",
"//tensorflow/models/research/object_detection/core:preprocessor",
"//tensorflow/models/research/object_detection/protos:image_resizer_py_pb2",
],
)
py_test(
name = "image_resizer_builder_test",
srcs = ["image_resizer_builder_test.py"],
deps = [
":image_resizer_builder",
"//tensorflow",
"//tensorflow/models/research/object_detection/protos:image_resizer_py_pb2",
],
)
research/object_detection/builders/anchor_generator_builder.py
View file @ 78d5f8f8
......@@ -87,7 +87,8 @@ def build(anchor_generator_config):
cfg.max_level,
cfg.anchor_scale,
[float(aspect_ratio) for aspect_ratio in cfg.aspect_ratios],
cfg.scales_per_octave
cfg.scales_per_octave,
cfg.normalize_coordinates
)
else:
raise ValueError('Empty anchor generator.')
research/object_detection/builders/anchor_generator_builder_test.py
View file @ 78d5f8f8
......@@ -276,6 +276,24 @@ class AnchorGeneratorBuilderTest(tf.test.TestCase):
self.assertAllClose(anchor_grid_info['info'][2],
[4.0 * 2**level, 4.0 * 2**level])
self.assertAllClose(anchor_grid_info['info'][3], [2**level, 2**level])
self.assertTrue(anchor_generator_object._normalize_coordinates)
def test_build_multiscale_anchor_generator_with_anchors_in_pixel_coordinates(
self):
anchor_generator_text_proto = """
multiscale_anchor_generator {
aspect_ratios: [1.0]
normalize_coordinates: false
}
"""
anchor_generator_proto = anchor_generator_pb2.AnchorGenerator()
text_format.Merge(anchor_generator_text_proto, anchor_generator_proto)
anchor_generator_object = anchor_generator_builder.build(
anchor_generator_proto)
self.assertTrue(isinstance(anchor_generator_object,
multiscale_grid_anchor_generator.
MultiscaleGridAnchorGenerator))
self.assertFalse(anchor_generator_object._normalize_coordinates)
if __name__ == '__main__':
......
research/object_detection/builders/hyperparams_builder.py
View file @ 78d5f8f8
......@@ -134,6 +134,10 @@ def _build_initializer(initializer):
return tf.truncated_normal_initializer(
mean=initializer.truncated_normal_initializer.mean,
stddev=initializer.truncated_normal_initializer.stddev)
if initializer_oneof == 'random_normal_initializer':
return tf.random_normal_initializer(
mean=initializer.random_normal_initializer.mean,
stddev=initializer.random_normal_initializer.stddev)
if initializer_oneof == 'variance_scaling_initializer':
enum_descriptor = (hyperparams_pb2.VarianceScalingInitializer.
DESCRIPTOR.enum_types_by_name['Mode'])
......
research/object_detection/builders/hyperparams_builder_test.py
View file @ 78d5f8f8
......@@ -28,7 +28,7 @@ slim = tf.contrib.slim
class HyperparamsBuilderTest(tf.test.TestCase):
# TODO: Make this a public api in slim arg_scope.py.
# TODO(rathodv): Make this a public api in slim arg_scope.py.
def _get_scope_key(self, op):
return getattr(op, '_key_op', str(op))
......@@ -444,6 +444,27 @@ class HyperparamsBuilderTest(tf.test.TestCase):
self._assert_variance_in_range(initializer, shape=[100, 40],
variance=0.49, tol=1e-1)
def test_variance_in_range_with_random_normal_initializer(self):
conv_hyperparams_text_proto = """
regularizer {
l2_regularizer {
}
}
initializer {
random_normal_initializer {
mean: 0.0
stddev: 0.8
}
}
"""
conv_hyperparams_proto = hyperparams_pb2.Hyperparams()
text_format.Merge(conv_hyperparams_text_proto, conv_hyperparams_proto)
scope = hyperparams_builder.build(conv_hyperparams_proto, is_training=True)
conv_scope_arguments = scope.values()[0]
initializer = conv_scope_arguments['weights_initializer']
self._assert_variance_in_range(initializer, shape=[100, 40],
variance=0.64, tol=1e-1)
if __name__ == '__main__':
tf.test.main()
research/object_detection/builders/image_resizer_builder.py
View file @ 78d5f8f8
......@@ -72,8 +72,8 @@ def build(image_resizer_config):
raise ValueError('image_resizer_config not of type '
'image_resizer_pb2.ImageResizer.')
if image_resizer_config.WhichOneof(
'image_resizer_oneof') == 'keep_aspect_ratio_resizer':
image_resizer_oneof = image_resizer_config.WhichOneof('image_resizer_oneof')
if image_resizer_oneof == 'keep_aspect_ratio_resizer':
keep_aspect_ratio_config = image_resizer_config.keep_aspect_ratio_resizer
if not (keep_aspect_ratio_config.min_dimension <=
keep_aspect_ratio_config.max_dimension):
......@@ -87,8 +87,7 @@ def build(image_resizer_config):
pad_to_max_dimension=keep_aspect_ratio_config.pad_to_max_dimension)
if not keep_aspect_ratio_config.convert_to_grayscale:
return image_resizer_fn
elif image_resizer_config.WhichOneof(
'image_resizer_oneof') == 'fixed_shape_resizer':
elif image_resizer_oneof == 'fixed_shape_resizer':
fixed_shape_resizer_config = image_resizer_config.fixed_shape_resizer
method = _tf_resize_method(fixed_shape_resizer_config.resize_method)
image_resizer_fn = functools.partial(
......@@ -99,7 +98,8 @@ def build(image_resizer_config):
if not fixed_shape_resizer_config.convert_to_grayscale:
return image_resizer_fn
else:
raise ValueError('Invalid image resizer option.')
raise ValueError(
'Invalid image resizer option: \'%s\'.' % image_resizer_oneof)
def grayscale_image_resizer(image):
[resized_image, resized_image_shape] = image_resizer_fn(image)
......
research/object_detection/builders/losses_builder.py
View file @ 78d5f8f8
......@@ -150,7 +150,8 @@ def _build_localization_loss(loss_config):
return losses.WeightedL2LocalizationLoss()
if loss_type == 'weighted_smooth_l1':
return losses.WeightedSmoothL1LocalizationLoss()
return losses.WeightedSmoothL1LocalizationLoss(
loss_config.weighted_smooth_l1.delta)
if loss_type == 'weighted_iou':
return losses.WeightedIOULocalizationLoss()
......
research/object_detection/builders/losses_builder_test.py
View file @ 78d5f8f8
......@@ -42,7 +42,7 @@ class LocalizationLossBuilderTest(tf.test.TestCase):
self.assertTrue(isinstance(localization_loss,
losses.WeightedL2LocalizationLoss))
def test_build_weighted_smooth_l1_localization_loss(self):
def test_build_weighted_smooth_l1_localization_loss_default_delta(self):
losses_text_proto = """
localization_loss {
weighted_smooth_l1 {
......@@ -58,6 +58,26 @@ class LocalizationLossBuilderTest(tf.test.TestCase):
_, localization_loss, _, _, _ = losses_builder.build(losses_proto)
self.assertTrue(isinstance(localization_loss,
losses.WeightedSmoothL1LocalizationLoss))
self.assertAlmostEqual(localization_loss._delta, 1.0)
def test_build_weighted_smooth_l1_localization_loss_non_default_delta(self):
losses_text_proto = """
localization_loss {
weighted_smooth_l1 {
delta: 0.1
}
}
classification_loss {
weighted_softmax {
}
}
"""
losses_proto = losses_pb2.Loss()
text_format.Merge(losses_text_proto, losses_proto)
_, localization_loss, _, _, _ = losses_builder.build(losses_proto)
self.assertTrue(isinstance(localization_loss,
losses.WeightedSmoothL1LocalizationLoss))
self.assertAlmostEqual(localization_loss._delta, 0.1)
def test_build_weighted_iou_localization_loss(self):
losses_text_proto = """
......
research/object_detection/builders/model_builder.py
View file @ 78d5f8f8
......@@ -153,6 +153,7 @@ def _build_ssd_model(ssd_config, is_training, add_summaries):
region_similarity_calculator = sim_calc.build(
ssd_config.similarity_calculator)
encode_background_as_zeros = ssd_config.encode_background_as_zeros
negative_class_weight = ssd_config.negative_class_weight
ssd_box_predictor = box_predictor_builder.build(hyperparams_builder.build,
ssd_config.box_predictor,
is_training, num_classes)
......@@ -165,6 +166,7 @@ def _build_ssd_model(ssd_config, is_training, add_summaries):
localization_weight,
hard_example_miner) = losses_builder.build(ssd_config.loss)
normalize_loss_by_num_matches = ssd_config.normalize_loss_by_num_matches
normalize_loc_loss_by_codesize = ssd_config.normalize_loc_loss_by_codesize
return ssd_meta_arch.SSDMetaArch(
is_training,
......@@ -175,6 +177,7 @@ def _build_ssd_model(ssd_config, is_training, add_summaries):
matcher,
region_similarity_calculator,
encode_background_as_zeros,
negative_class_weight,
image_resizer_fn,
non_max_suppression_fn,
score_conversion_fn,
......@@ -184,7 +187,8 @@ def _build_ssd_model(ssd_config, is_training, add_summaries):
localization_weight,
normalize_loss_by_num_matches,
hard_example_miner,
add_summaries=add_summaries)
add_summaries=add_summaries,
normalize_loc_loss_by_codesize=normalize_loc_loss_by_codesize)
def _build_faster_rcnn_feature_extractor(
......
research/object_detection/builders/model_builder_test.py
View file @ 78d5f8f8
......@@ -259,13 +259,15 @@ class ModelBuilderTest(tf.test.TestCase):
}
}
initializer {
truncated_normal_initializer {
random_normal_initializer {
}
}
}
num_layers_before_predictor: 1
}
}
normalize_loss_by_num_matches: true
normalize_loc_loss_by_codesize: true
loss {
classification_loss {
weighted_sigmoid_focal {
......@@ -275,6 +277,7 @@ class ModelBuilderTest(tf.test.TestCase):
}
localization_loss {
weighted_smooth_l1 {
delta: 0.1
}
}
classification_weight: 1.0
......@@ -344,6 +347,7 @@ class ModelBuilderTest(tf.test.TestCase):
}
}
}
normalize_loc_loss_by_codesize: true
loss {
classification_loss {
weighted_softmax {
......@@ -362,6 +366,7 @@ class ModelBuilderTest(tf.test.TestCase):
self.assertIsInstance(model._feature_extractor,
SSDMobileNetV1FeatureExtractor)
self.assertTrue(model._feature_extractor._batch_norm_trainable)
self.assertTrue(model._normalize_loc_loss_by_codesize)
def test_create_embedded_ssd_mobilenet_v1_model_from_config(self):
model_text_proto = """
......
research/object_detection/builders/optimizer_builder_test.py
View file @ 78d5f8f8
......@@ -187,7 +187,7 @@ class OptimizerBuilderTest(tf.test.TestCase):
optimizer, _ = optimizer_builder.build(optimizer_proto)
self.assertTrue(
isinstance(optimizer, tf.contrib.opt.MovingAverageOptimizer))
# TODO: Find a way to not depend on the private members.
# TODO(rathodv): Find a way to not depend on the private members.
self.assertAlmostEqual(optimizer._ema._decay, 0.2)
def testBuildEmptyOptimizer(self):
......
Markdown is supported
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment