Skip to content

GitLab

Commit a4944a57 authored by derekjchow's avatar derekjchow Committed by Sergio Guadarrama
Browse files

Add Tensorflow Object Detection API. (#1561) 

For details see our paper:
"Speed/accuracy trade-offs for modern convolutional object detectors."
Huang J, Rathod V, Sun C, Zhu M, Korattikara A, Fathi A, Fischer I,
Wojna Z, Song Y, Guadarrama S, Murphy K, CVPR 2017
https://arxiv.org/abs/1611.10012
parent 60c3ed2e
Hide whitespace changes
Inline Side-by-side
Showing with 3514 additions and 0 deletions
object_detection/builders/BUILD 0 → 100644
View file @ a4944a57
# 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/object_detection/core:box_predictor",
"//tensorflow_models/object_detection/meta_architectures:faster_rcnn_meta_arch",
"//tensorflow_models/object_detection/meta_architectures:rfcn_meta_arch",
"//tensorflow_models/object_detection/meta_architectures:ssd_meta_arch",
"//tensorflow_models/object_detection/models:faster_rcnn_inception_resnet_v2_feature_extractor",
"//tensorflow_models/object_detection/models:faster_rcnn_resnet_v1_feature_extractor",
"//tensorflow_models/object_detection/models:ssd_inception_v2_feature_extractor",
"//tensorflow_models/object_detection/models:ssd_mobilenet_v1_feature_extractor",
"//tensorflow_models/object_detection/protos:model_py_pb2",
],
)
py_test(
name = "model_builder_test",
srcs = ["model_builder_test.py"],
deps = [
":model_builder",
"//tensorflow",
"//tensorflow_models/object_detection/meta_architectures:faster_rcnn_meta_arch",
"//tensorflow_models/object_detection/meta_architectures:ssd_meta_arch",
"//tensorflow_models/object_detection/models:ssd_inception_v2_feature_extractor",
"//tensorflow_models/object_detection/models:ssd_mobilenet_v1_feature_extractor",
"//tensorflow_models/object_detection/protos:model_py_pb2",
],
)
py_library(
name = "matcher_builder",
srcs = ["matcher_builder.py"],
deps = [
"//tensorflow_models/object_detection/matchers:argmax_matcher",
"//tensorflow_models/object_detection/matchers:bipartite_matcher",
"//tensorflow_models/object_detection/protos:matcher_py_pb2",
],
)
py_test(
name = "matcher_builder_test",
srcs = ["matcher_builder_test.py"],
deps = [
":matcher_builder",
"//tensorflow_models/object_detection/matchers:argmax_matcher",
"//tensorflow_models/object_detection/matchers:bipartite_matcher",
"//tensorflow_models/object_detection/protos:matcher_py_pb2",
],
)
py_library(
name = "box_coder_builder",
srcs = ["box_coder_builder.py"],
deps = [
"//tensorflow_models/object_detection/box_coders:faster_rcnn_box_coder",
"//tensorflow_models/object_detection/box_coders:mean_stddev_box_coder",
"//tensorflow_models/object_detection/box_coders:square_box_coder",
"//tensorflow_models/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/object_detection/box_coders:faster_rcnn_box_coder",
"//tensorflow_models/object_detection/box_coders:mean_stddev_box_coder",
"//tensorflow_models/object_detection/box_coders:square_box_coder",
"//tensorflow_models/object_detection/protos:box_coder_py_pb2",
],
)
py_library(
name = "anchor_generator_builder",
srcs = ["anchor_generator_builder.py"],
deps = [
"//tensorflow_models/object_detection/anchor_generators:grid_anchor_generator",
"//tensorflow_models/object_detection/anchor_generators:multiple_grid_anchor_generator",
"//tensorflow_models/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/object_detection/anchor_generators:grid_anchor_generator",
"//tensorflow_models/object_detection/anchor_generators:multiple_grid_anchor_generator",
"//tensorflow_models/object_detection/protos:anchor_generator_py_pb2",
],
)
py_library(
name = "input_reader_builder",
srcs = ["input_reader_builder.py"],
deps = [
"//tensorflow",
"//tensorflow_models/object_detection/data_decoders:tf_example_decoder",
"//tensorflow_models/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/object_detection/core:standard_fields",
"//tensorflow_models/object_detection/protos:input_reader_py_pb2",
],
)
py_library(
name = "losses_builder",
srcs = ["losses_builder.py"],
deps = [
"//tensorflow_models/object_detection/core:losses",
"//tensorflow_models/object_detection/protos:losses_py_pb2",
],
)
py_test(
name = "losses_builder_test",
srcs = ["losses_builder_test.py"],
deps = [
":losses_builder",
"//tensorflow_models/object_detection/core:losses",
"//tensorflow_models/object_detection/protos:losses_py_pb2",
],
)
py_library(
name = "optimizer_builder",
srcs = ["optimizer_builder.py"],
deps = [
"//tensorflow",
"//tensorflow_models/object_detection/utils:learning_schedules",
],
)
py_test(
name = "optimizer_builder_test",
srcs = ["optimizer_builder_test.py"],
deps = [
":optimizer_builder",
"//tensorflow",
"//tensorflow_models/object_detection/protos:optimizer_py_pb2",
],
)
py_library(
name = "post_processing_builder",
srcs = ["post_processing_builder.py"],
deps = [
"//tensorflow",
"//tensorflow_models/object_detection/core:post_processing",
"//tensorflow_models/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/object_detection/protos:post_processing_py_pb2",
],
)
py_library(
name = "hyperparams_builder",
srcs = ["hyperparams_builder.py"],
deps = [
"//tensorflow_models/object_detection/protos:hyperparams_py_pb2",
],
)
py_test(
name = "hyperparams_builder_test",
srcs = ["hyperparams_builder_test.py"],
deps = [
":hyperparams_builder",
"//tensorflow",
"//tensorflow_models/object_detection/protos:hyperparams_py_pb2",
],
)
py_library(
name = "box_predictor_builder",
srcs = ["box_predictor_builder.py"],
deps = [
":hyperparams_builder",
"//tensorflow_models/object_detection/core:box_predictor",
"//tensorflow_models/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/object_detection/protos:box_predictor_py_pb2",
"//tensorflow_models/object_detection/protos:hyperparams_py_pb2",
],
)
py_library(
name = "region_similarity_calculator_builder",
srcs = ["region_similarity_calculator_builder.py"],
deps = [
"//tensorflow_models/object_detection/core:region_similarity_calculator",
"//tensorflow_models/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/object_detection/core:preprocessor",
"//tensorflow_models/object_detection/protos:preprocessor_py_pb2",
],
)
py_test(
name = "preprocessor_builder_test",
srcs = [
"preprocessor_builder_test.py",
],
deps = [
":preprocessor_builder",
"//tensorflow",
"//tensorflow_models/object_detection/core:preprocessor",
"//tensorflow_models/object_detection/protos:preprocessor_py_pb2",
],
)
py_library(
name = "image_resizer_builder",
srcs = ["image_resizer_builder.py"],
deps = [
"//tensorflow",
"//tensorflow_models/object_detection/core:preprocessor",
"//tensorflow_models/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/object_detection/protos:image_resizer_py_pb2",
],
)
object_detection/builders/anchor_generator_builder.py 0 → 100644
View file @ a4944a57
# Copyright 2017 The TensorFlow Authors. All Rights Reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
# ==============================================================================
"""A function to build an object detection anchor generator from config."""
from object_detection.anchor_generators import grid_anchor_generator
from object_detection.anchor_generators import multiple_grid_anchor_generator
from object_detection.protos import anchor_generator_pb2
def build(anchor_generator_config):
"""Builds an anchor generator based on the config.
Args:
anchor_generator_config: An anchor_generator.proto object containing the
config for the desired anchor generator.
Returns:
Anchor generator based on the config.
Raises:
ValueError: On empty anchor generator proto.
"""
if not isinstance(anchor_generator_config,
anchor_generator_pb2.AnchorGenerator):
raise ValueError('anchor_generator_config not of type '
'anchor_generator_pb2.AnchorGenerator')
if anchor_generator_config.WhichOneof(
'anchor_generator_oneof') == 'grid_anchor_generator':
grid_anchor_generator_config = anchor_generator_config.grid_anchor_generator
return grid_anchor_generator.GridAnchorGenerator(
scales=[float(scale) for scale in grid_anchor_generator_config.scales],
aspect_ratios=[float(aspect_ratio)
for aspect_ratio
in grid_anchor_generator_config.aspect_ratios],
base_anchor_size=[grid_anchor_generator_config.height,
grid_anchor_generator_config.width],
anchor_stride=[grid_anchor_generator_config.height_stride,
grid_anchor_generator_config.width_stride],
anchor_offset=[grid_anchor_generator_config.height_offset,
grid_anchor_generator_config.width_offset])
elif anchor_generator_config.WhichOneof(
'anchor_generator_oneof') == 'ssd_anchor_generator':
ssd_anchor_generator_config = anchor_generator_config.ssd_anchor_generator
return multiple_grid_anchor_generator.create_ssd_anchors(
num_layers=ssd_anchor_generator_config.num_layers,
min_scale=ssd_anchor_generator_config.min_scale,
max_scale=ssd_anchor_generator_config.max_scale,
aspect_ratios=ssd_anchor_generator_config.aspect_ratios,
reduce_boxes_in_lowest_layer=(ssd_anchor_generator_config
.reduce_boxes_in_lowest_layer))
else:
raise ValueError('Empty anchor generator.')
object_detection/builders/anchor_generator_builder_test.py 0 → 100644
View file @ a4944a57
# Copyright 2017 The TensorFlow Authors. All Rights Reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
# ==============================================================================
"""Tests for anchor_generator_builder."""
import tensorflow as tf
from google.protobuf import text_format
from object_detection.anchor_generators import grid_anchor_generator
from object_detection.anchor_generators import multiple_grid_anchor_generator
from object_detection.builders import anchor_generator_builder
from object_detection.protos import anchor_generator_pb2
class AnchorGeneratorBuilderTest(tf.test.TestCase):
def assert_almost_list_equal(self, expected_list, actual_list, delta=None):
self.assertEqual(len(expected_list), len(actual_list))
for expected_item, actual_item in zip(expected_list, actual_list):
self.assertAlmostEqual(expected_item, actual_item, delta=delta)
def test_build_grid_anchor_generator_with_defaults(self):
anchor_generator_text_proto = """
grid_anchor_generator {
}
"""
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,
grid_anchor_generator.GridAnchorGenerator))
self.assertListEqual(anchor_generator_object._scales, [])
self.assertListEqual(anchor_generator_object._aspect_ratios, [])
with self.test_session() as sess:
base_anchor_size, anchor_offset, anchor_stride = sess.run(
[anchor_generator_object._base_anchor_size,
anchor_generator_object._anchor_offset,
anchor_generator_object._anchor_stride])
self.assertAllEqual(anchor_offset, [0, 0])
self.assertAllEqual(anchor_stride, [16, 16])
self.assertAllEqual(base_anchor_size, [256, 256])
def test_build_grid_anchor_generator_with_non_default_parameters(self):
anchor_generator_text_proto = """
grid_anchor_generator {
height: 128
width: 512
height_stride: 10
width_stride: 20
height_offset: 30
width_offset: 40
scales: [0.4, 2.2]
aspect_ratios: [0.3, 4.5]
}
"""
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,
grid_anchor_generator.GridAnchorGenerator))
self.assert_almost_list_equal(anchor_generator_object._scales,
[0.4, 2.2])
self.assert_almost_list_equal(anchor_generator_object._aspect_ratios,
[0.3, 4.5])
with self.test_session() as sess:
base_anchor_size, anchor_offset, anchor_stride = sess.run(
[anchor_generator_object._base_anchor_size,
anchor_generator_object._anchor_offset,
anchor_generator_object._anchor_stride])
self.assertAllEqual(anchor_offset, [30, 40])
self.assertAllEqual(anchor_stride, [10, 20])
self.assertAllEqual(base_anchor_size, [128, 512])
def test_build_ssd_anchor_generator_with_defaults(self):
anchor_generator_text_proto = """
ssd_anchor_generator {
aspect_ratios: [1.0]
}
"""
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,
multiple_grid_anchor_generator.
MultipleGridAnchorGenerator))
for actual_scales, expected_scales in zip(
list(anchor_generator_object._scales),
[(0.1, 0.2, 0.2),
(0.35, 0.418),
(0.499, 0.570),
(0.649, 0.721),
(0.799, 0.871),
(0.949, 0.974)]):
self.assert_almost_list_equal(expected_scales, actual_scales, delta=1e-2)
for actual_aspect_ratio, expected_aspect_ratio in zip(
list(anchor_generator_object._aspect_ratios),
[(1.0, 2.0, 0.5)] + 5 * [(1.0, 1.0)]):
self.assert_almost_list_equal(expected_aspect_ratio, actual_aspect_ratio)
with self.test_session() as sess:
base_anchor_size = sess.run(anchor_generator_object._base_anchor_size)
self.assertAllClose(base_anchor_size, [1.0, 1.0])
def test_build_ssd_anchor_generator_withoud_reduced_boxes(self):
anchor_generator_text_proto = """
ssd_anchor_generator {
aspect_ratios: [1.0]
reduce_boxes_in_lowest_layer: 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,
multiple_grid_anchor_generator.
MultipleGridAnchorGenerator))
for actual_scales, expected_scales in zip(
list(anchor_generator_object._scales),
[(0.2, 0.264),
(0.35, 0.418),
(0.499, 0.570),
(0.649, 0.721),
(0.799, 0.871),
(0.949, 0.974)]):
self.assert_almost_list_equal(expected_scales, actual_scales, delta=1e-2)
for actual_aspect_ratio, expected_aspect_ratio in zip(
list(anchor_generator_object._aspect_ratios),
6 * [(1.0, 1.0)]):
self.assert_almost_list_equal(expected_aspect_ratio, actual_aspect_ratio)
with self.test_session() as sess:
base_anchor_size = sess.run(anchor_generator_object._base_anchor_size)
self.assertAllClose(base_anchor_size, [1.0, 1.0])
def test_build_ssd_anchor_generator_with_non_default_parameters(self):
anchor_generator_text_proto = """
ssd_anchor_generator {
num_layers: 2
min_scale: 0.3
max_scale: 0.8
aspect_ratios: [2.0]
}
"""
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,
multiple_grid_anchor_generator.
MultipleGridAnchorGenerator))
for actual_scales, expected_scales in zip(
list(anchor_generator_object._scales),
[(0.1, 0.3, 0.3), (0.8,)]):
self.assert_almost_list_equal(expected_scales, actual_scales, delta=1e-2)
for actual_aspect_ratio, expected_aspect_ratio in zip(
list(anchor_generator_object._aspect_ratios),
[(1.0, 2.0, 0.5), (2.0,)]):
self.assert_almost_list_equal(expected_aspect_ratio, actual_aspect_ratio)
with self.test_session() as sess:
base_anchor_size = sess.run(anchor_generator_object._base_anchor_size)
self.assertAllClose(base_anchor_size, [1.0, 1.0])
def test_raise_value_error_on_empty_anchor_genertor(self):
anchor_generator_text_proto = """
"""
anchor_generator_proto = anchor_generator_pb2.AnchorGenerator()
text_format.Merge(anchor_generator_text_proto, anchor_generator_proto)
with self.assertRaises(ValueError):
anchor_generator_builder.build(anchor_generator_proto)
if __name__ == '__main__':
tf.test.main()
object_detection/builders/box_coder_builder.py 0 → 100644
View file @ a4944a57
# Copyright 2017 The TensorFlow Authors. All Rights Reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
# ==============================================================================
"""A function to build an object detection box coder from configuration."""
from object_detection.box_coders import faster_rcnn_box_coder
from object_detection.box_coders import mean_stddev_box_coder
from object_detection.box_coders import square_box_coder
from object_detection.protos import box_coder_pb2
def build(box_coder_config):
"""Builds a box coder object based on the box coder config.
Args:
box_coder_config: A box_coder.proto object containing the config for the
desired box coder.
Returns:
BoxCoder based on the config.
Raises:
ValueError: On empty box coder proto.
"""
if not isinstance(box_coder_config, box_coder_pb2.BoxCoder):
raise ValueError('box_coder_config not of type box_coder_pb2.BoxCoder.')
if box_coder_config.WhichOneof('box_coder_oneof') == 'faster_rcnn_box_coder':
return faster_rcnn_box_coder.FasterRcnnBoxCoder(scale_factors=[
box_coder_config.faster_rcnn_box_coder.y_scale,
box_coder_config.faster_rcnn_box_coder.x_scale,
box_coder_config.faster_rcnn_box_coder.height_scale,
box_coder_config.faster_rcnn_box_coder.width_scale
])
if (box_coder_config.WhichOneof('box_coder_oneof') ==
'mean_stddev_box_coder'):
return mean_stddev_box_coder.MeanStddevBoxCoder()
if box_coder_config.WhichOneof('box_coder_oneof') == 'square_box_coder':
return square_box_coder.SquareBoxCoder(scale_factors=[
box_coder_config.square_box_coder.y_scale,
box_coder_config.square_box_coder.x_scale,
box_coder_config.square_box_coder.length_scale
])
raise ValueError('Empty box coder.')
object_detection/builders/box_coder_builder_test.py 0 → 100644
View file @ a4944a57
# Copyright 2017 The TensorFlow Authors. All Rights Reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
# ==============================================================================
"""Tests for box_coder_builder."""
import tensorflow as tf
from google.protobuf import text_format
from object_detection.box_coders import faster_rcnn_box_coder
from object_detection.box_coders import mean_stddev_box_coder
from object_detection.box_coders import square_box_coder
from object_detection.builders import box_coder_builder
from object_detection.protos import box_coder_pb2
class BoxCoderBuilderTest(tf.test.TestCase):
def test_build_faster_rcnn_box_coder_with_defaults(self):
box_coder_text_proto = """
faster_rcnn_box_coder {
}
"""
box_coder_proto = box_coder_pb2.BoxCoder()
text_format.Merge(box_coder_text_proto, box_coder_proto)
box_coder_object = box_coder_builder.build(box_coder_proto)
self.assertTrue(isinstance(box_coder_object,
faster_rcnn_box_coder.FasterRcnnBoxCoder))
self.assertEqual(box_coder_object._scale_factors, [10.0, 10.0, 5.0, 5.0])
def test_build_faster_rcnn_box_coder_with_non_default_parameters(self):
box_coder_text_proto = """
faster_rcnn_box_coder {
y_scale: 6.0
x_scale: 3.0
height_scale: 7.0
width_scale: 8.0
}
"""
box_coder_proto = box_coder_pb2.BoxCoder()
text_format.Merge(box_coder_text_proto, box_coder_proto)
box_coder_object = box_coder_builder.build(box_coder_proto)
self.assertTrue(isinstance(box_coder_object,
faster_rcnn_box_coder.FasterRcnnBoxCoder))
self.assertEqual(box_coder_object._scale_factors, [6.0, 3.0, 7.0, 8.0])
def test_build_mean_stddev_box_coder(self):
box_coder_text_proto = """
mean_stddev_box_coder {
}
"""
box_coder_proto = box_coder_pb2.BoxCoder()
text_format.Merge(box_coder_text_proto, box_coder_proto)
box_coder_object = box_coder_builder.build(box_coder_proto)
self.assertTrue(
isinstance(box_coder_object,
mean_stddev_box_coder.MeanStddevBoxCoder))
def test_build_square_box_coder_with_defaults(self):
box_coder_text_proto = """
square_box_coder {
}
"""
box_coder_proto = box_coder_pb2.BoxCoder()
text_format.Merge(box_coder_text_proto, box_coder_proto)
box_coder_object = box_coder_builder.build(box_coder_proto)
self.assertTrue(
isinstance(box_coder_object, square_box_coder.SquareBoxCoder))
self.assertEqual(box_coder_object._scale_factors, [10.0, 10.0, 5.0])
def test_build_square_box_coder_with_non_default_parameters(self):
box_coder_text_proto = """
square_box_coder {
y_scale: 6.0
x_scale: 3.0
length_scale: 7.0
}
"""
box_coder_proto = box_coder_pb2.BoxCoder()
text_format.Merge(box_coder_text_proto, box_coder_proto)
box_coder_object = box_coder_builder.build(box_coder_proto)
self.assertTrue(
isinstance(box_coder_object, square_box_coder.SquareBoxCoder))
self.assertEqual(box_coder_object._scale_factors, [6.0, 3.0, 7.0])
def test_raise_error_on_empty_box_coder(self):
box_coder_text_proto = """
"""
box_coder_proto = box_coder_pb2.BoxCoder()
text_format.Merge(box_coder_text_proto, box_coder_proto)
with self.assertRaises(ValueError):
box_coder_builder.build(box_coder_proto)
if __name__ == '__main__':
tf.test.main()
object_detection/builders/box_predictor_builder.py 0 → 100644
View file @ a4944a57
# Copyright 2017 The TensorFlow Authors. All Rights Reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
# ==============================================================================
"""Function to build box predictor from configuration."""
from object_detection.core import box_predictor
from object_detection.protos import box_predictor_pb2
def build(argscope_fn, box_predictor_config, is_training, num_classes):
"""Builds box predictor based on the configuration.
Builds box predictor based on the configuration. See box_predictor.proto for
configurable options. Also, see box_predictor.py for more details.
Args:
argscope_fn: A function that takes the following inputs:
* hyperparams_pb2.Hyperparams proto
* a boolean indicating if the model is in training mode.
and returns a tf slim argscope for Conv and FC hyperparameters.
box_predictor_config: box_predictor_pb2.BoxPredictor proto containing
configuration.
is_training: Whether the models is in training mode.
num_classes: Number of classes to predict.
Returns:
box_predictor: box_predictor.BoxPredictor object.
Raises:
ValueError: On unknown box predictor.
"""
if not isinstance(box_predictor_config, box_predictor_pb2.BoxPredictor):
raise ValueError('box_predictor_config not of type '
'box_predictor_pb2.BoxPredictor.')
box_predictor_oneof = box_predictor_config.WhichOneof('box_predictor_oneof')
if box_predictor_oneof == 'convolutional_box_predictor':
conv_box_predictor = box_predictor_config.convolutional_box_predictor
conv_hyperparams = argscope_fn(conv_box_predictor.conv_hyperparams,
is_training)
box_predictor_object = box_predictor.ConvolutionalBoxPredictor(
is_training=is_training,
num_classes=num_classes,
conv_hyperparams=conv_hyperparams,
min_depth=conv_box_predictor.min_depth,
max_depth=conv_box_predictor.max_depth,
num_layers_before_predictor=(conv_box_predictor.
num_layers_before_predictor),
use_dropout=conv_box_predictor.use_dropout,
dropout_keep_prob=conv_box_predictor.dropout_keep_probability,
kernel_size=conv_box_predictor.kernel_size,
box_code_size=conv_box_predictor.box_code_size,
apply_sigmoid_to_scores=conv_box_predictor.apply_sigmoid_to_scores)
return box_predictor_object
if box_predictor_oneof == 'mask_rcnn_box_predictor':
mask_rcnn_box_predictor = box_predictor_config.mask_rcnn_box_predictor
fc_hyperparams = argscope_fn(mask_rcnn_box_predictor.fc_hyperparams,
is_training)
conv_hyperparams = None
if mask_rcnn_box_predictor.HasField('conv_hyperparams'):
conv_hyperparams = argscope_fn(mask_rcnn_box_predictor.conv_hyperparams,
is_training)
box_predictor_object = box_predictor.MaskRCNNBoxPredictor(
is_training=is_training,
num_classes=num_classes,
fc_hyperparams=fc_hyperparams,
use_dropout=mask_rcnn_box_predictor.use_dropout,
dropout_keep_prob=mask_rcnn_box_predictor.dropout_keep_probability,
box_code_size=mask_rcnn_box_predictor.box_code_size,
conv_hyperparams=conv_hyperparams,
predict_instance_masks=mask_rcnn_box_predictor.predict_instance_masks,
mask_prediction_conv_depth=(mask_rcnn_box_predictor.
mask_prediction_conv_depth),
predict_keypoints=mask_rcnn_box_predictor.predict_keypoints)
return box_predictor_object
if box_predictor_oneof == 'rfcn_box_predictor':
rfcn_box_predictor = box_predictor_config.rfcn_box_predictor
conv_hyperparams = argscope_fn(rfcn_box_predictor.conv_hyperparams,
is_training)
box_predictor_object = box_predictor.RfcnBoxPredictor(
is_training=is_training,
num_classes=num_classes,
conv_hyperparams=conv_hyperparams,
crop_size=[rfcn_box_predictor.crop_height,
rfcn_box_predictor.crop_width],
num_spatial_bins=[rfcn_box_predictor.num_spatial_bins_height,
rfcn_box_predictor.num_spatial_bins_width],
depth=rfcn_box_predictor.depth,
box_code_size=rfcn_box_predictor.box_code_size)
return box_predictor_object
raise ValueError('Unknown box predictor: {}'.format(box_predictor_oneof))
object_detection/builders/box_predictor_builder_test.py 0 → 100644
View file @ a4944a57
# Copyright 2017 The TensorFlow Authors. All Rights Reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
# ==============================================================================
"""Tests for box_predictor_builder."""
import mock
import tensorflow as tf
from google.protobuf import text_format
from object_detection.builders import box_predictor_builder
from object_detection.builders import hyperparams_builder
from object_detection.protos import box_predictor_pb2
from object_detection.protos import hyperparams_pb2
class ConvolutionalBoxPredictorBuilderTest(tf.test.TestCase):
def test_box_predictor_calls_conv_argscope_fn(self):
conv_hyperparams_text_proto = """
regularizer {
l1_regularizer {
weight: 0.0003
}
}
initializer {
truncated_normal_initializer {
mean: 0.0
stddev: 0.3
}
}
activation: RELU_6
"""
hyperparams_proto = hyperparams_pb2.Hyperparams()
text_format.Merge(conv_hyperparams_text_proto, hyperparams_proto)
def mock_conv_argscope_builder(conv_hyperparams_arg, is_training):
return (conv_hyperparams_arg, is_training)
box_predictor_proto = box_predictor_pb2.BoxPredictor()
box_predictor_proto.convolutional_box_predictor.conv_hyperparams.CopyFrom(
hyperparams_proto)
box_predictor = box_predictor_builder.build(
argscope_fn=mock_conv_argscope_builder,
box_predictor_config=box_predictor_proto,
is_training=False,
num_classes=10)
(conv_hyperparams_actual, is_training) = box_predictor._conv_hyperparams
self.assertAlmostEqual((hyperparams_proto.regularizer.
l1_regularizer.weight),
(conv_hyperparams_actual.regularizer.l1_regularizer.
weight))
self.assertAlmostEqual((hyperparams_proto.initializer.
truncated_normal_initializer.stddev),
(conv_hyperparams_actual.initializer.
truncated_normal_initializer.stddev))
self.assertAlmostEqual((hyperparams_proto.initializer.
truncated_normal_initializer.mean),
(conv_hyperparams_actual.initializer.
truncated_normal_initializer.mean))
self.assertEqual(hyperparams_proto.activation,
conv_hyperparams_actual.activation)
self.assertFalse(is_training)
def test_construct_non_default_conv_box_predictor(self):
box_predictor_text_proto = """
convolutional_box_predictor {
min_depth: 2
max_depth: 16
num_layers_before_predictor: 2
use_dropout: false
dropout_keep_probability: 0.4
kernel_size: 3
box_code_size: 3
apply_sigmoid_to_scores: true
}
"""
conv_hyperparams_text_proto = """
regularizer {
l1_regularizer {
}
}
initializer {
truncated_normal_initializer {
}
}
"""
hyperparams_proto = hyperparams_pb2.Hyperparams()
text_format.Merge(conv_hyperparams_text_proto, hyperparams_proto)
def mock_conv_argscope_builder(conv_hyperparams_arg, is_training):
return (conv_hyperparams_arg, is_training)
box_predictor_proto = box_predictor_pb2.BoxPredictor()
text_format.Merge(box_predictor_text_proto, box_predictor_proto)
box_predictor_proto.convolutional_box_predictor.conv_hyperparams.CopyFrom(
hyperparams_proto)
box_predictor = box_predictor_builder.build(
argscope_fn=mock_conv_argscope_builder,
box_predictor_config=box_predictor_proto,
is_training=False,
num_classes=10)
self.assertEqual(box_predictor._min_depth, 2)
self.assertEqual(box_predictor._max_depth, 16)
self.assertEqual(box_predictor._num_layers_before_predictor, 2)
self.assertFalse(box_predictor._use_dropout)
self.assertAlmostEqual(box_predictor._dropout_keep_prob, 0.4)
self.assertTrue(box_predictor._apply_sigmoid_to_scores)
self.assertEqual(box_predictor.num_classes, 10)
self.assertFalse(box_predictor._is_training)
def test_construct_default_conv_box_predictor(self):
box_predictor_text_proto = """
convolutional_box_predictor {
conv_hyperparams {
regularizer {
l1_regularizer {
}
}
initializer {
truncated_normal_initializer {
}
}
}
}"""
box_predictor_proto = box_predictor_pb2.BoxPredictor()
text_format.Merge(box_predictor_text_proto, box_predictor_proto)
box_predictor = box_predictor_builder.build(
argscope_fn=hyperparams_builder.build,
box_predictor_config=box_predictor_proto,
is_training=True,
num_classes=90)
self.assertEqual(box_predictor._min_depth, 0)
self.assertEqual(box_predictor._max_depth, 0)
self.assertEqual(box_predictor._num_layers_before_predictor, 0)
self.assertTrue(box_predictor._use_dropout)
self.assertAlmostEqual(box_predictor._dropout_keep_prob, 0.8)
self.assertFalse(box_predictor._apply_sigmoid_to_scores)
self.assertEqual(box_predictor.num_classes, 90)
self.assertTrue(box_predictor._is_training)
class MaskRCNNBoxPredictorBuilderTest(tf.test.TestCase):
def test_box_predictor_builder_calls_fc_argscope_fn(self):
fc_hyperparams_text_proto = """
regularizer {
l1_regularizer {
weight: 0.0003
}
}
initializer {
truncated_normal_initializer {
mean: 0.0
stddev: 0.3
}
}
activation: RELU_6
op: FC
"""
hyperparams_proto = hyperparams_pb2.Hyperparams()
text_format.Merge(fc_hyperparams_text_proto, hyperparams_proto)
box_predictor_proto = box_predictor_pb2.BoxPredictor()
box_predictor_proto.mask_rcnn_box_predictor.fc_hyperparams.CopyFrom(
hyperparams_proto)
mock_argscope_fn = mock.Mock(return_value='arg_scope')
box_predictor = box_predictor_builder.build(
argscope_fn=mock_argscope_fn,
box_predictor_config=box_predictor_proto,
is_training=False,
num_classes=10)
mock_argscope_fn.assert_called_with(hyperparams_proto, False)
self.assertEqual(box_predictor._fc_hyperparams, 'arg_scope')
def test_non_default_mask_rcnn_box_predictor(self):
fc_hyperparams_text_proto = """
regularizer {
l1_regularizer {
}
}
initializer {
truncated_normal_initializer {
}
}
activation: RELU_6
op: FC
"""
box_predictor_text_proto = """
mask_rcnn_box_predictor {
use_dropout: true
dropout_keep_probability: 0.8
box_code_size: 3
}
"""
hyperparams_proto = hyperparams_pb2.Hyperparams()
text_format.Merge(fc_hyperparams_text_proto, hyperparams_proto)
def mock_fc_argscope_builder(fc_hyperparams_arg, is_training):
return (fc_hyperparams_arg, is_training)
box_predictor_proto = box_predictor_pb2.BoxPredictor()
text_format.Merge(box_predictor_text_proto, box_predictor_proto)
box_predictor_proto.mask_rcnn_box_predictor.fc_hyperparams.CopyFrom(
hyperparams_proto)
box_predictor = box_predictor_builder.build(
argscope_fn=mock_fc_argscope_builder,
box_predictor_config=box_predictor_proto,
is_training=True,
num_classes=90)
self.assertTrue(box_predictor._use_dropout)
self.assertAlmostEqual(box_predictor._dropout_keep_prob, 0.8)
self.assertEqual(box_predictor.num_classes, 90)
self.assertTrue(box_predictor._is_training)
self.assertEqual(box_predictor._box_code_size, 3)
def test_build_default_mask_rcnn_box_predictor(self):
box_predictor_proto = box_predictor_pb2.BoxPredictor()
box_predictor_proto.mask_rcnn_box_predictor.fc_hyperparams.op = (
hyperparams_pb2.Hyperparams.FC)
box_predictor = box_predictor_builder.build(
argscope_fn=mock.Mock(return_value='arg_scope'),
box_predictor_config=box_predictor_proto,
is_training=True,
num_classes=90)
self.assertFalse(box_predictor._use_dropout)
self.assertAlmostEqual(box_predictor._dropout_keep_prob, 0.5)
self.assertEqual(box_predictor.num_classes, 90)
self.assertTrue(box_predictor._is_training)
self.assertEqual(box_predictor._box_code_size, 4)
self.assertFalse(box_predictor._predict_instance_masks)
self.assertFalse(box_predictor._predict_keypoints)
def test_build_box_predictor_with_mask_branch(self):
box_predictor_proto = box_predictor_pb2.BoxPredictor()
box_predictor_proto.mask_rcnn_box_predictor.fc_hyperparams.op = (
hyperparams_pb2.Hyperparams.FC)
box_predictor_proto.mask_rcnn_box_predictor.conv_hyperparams.op = (
hyperparams_pb2.Hyperparams.CONV)
box_predictor_proto.mask_rcnn_box_predictor.predict_instance_masks = True
box_predictor_proto.mask_rcnn_box_predictor.mask_prediction_conv_depth = 512
mock_argscope_fn = mock.Mock(return_value='arg_scope')
box_predictor = box_predictor_builder.build(
argscope_fn=mock_argscope_fn,
box_predictor_config=box_predictor_proto,
is_training=True,
num_classes=90)
mock_argscope_fn.assert_has_calls(
[mock.call(box_predictor_proto.mask_rcnn_box_predictor.fc_hyperparams,
True),
mock.call(box_predictor_proto.mask_rcnn_box_predictor.conv_hyperparams,
True)], any_order=True)
self.assertFalse(box_predictor._use_dropout)
self.assertAlmostEqual(box_predictor._dropout_keep_prob, 0.5)
self.assertEqual(box_predictor.num_classes, 90)
self.assertTrue(box_predictor._is_training)
self.assertEqual(box_predictor._box_code_size, 4)
self.assertTrue(box_predictor._predict_instance_masks)
self.assertEqual(box_predictor._mask_prediction_conv_depth, 512)
self.assertFalse(box_predictor._predict_keypoints)
class RfcnBoxPredictorBuilderTest(tf.test.TestCase):
def test_box_predictor_calls_fc_argscope_fn(self):
conv_hyperparams_text_proto = """
regularizer {
l1_regularizer {
weight: 0.0003
}
}
initializer {
truncated_normal_initializer {
mean: 0.0
stddev: 0.3
}
}
activation: RELU_6
"""
hyperparams_proto = hyperparams_pb2.Hyperparams()
text_format.Merge(conv_hyperparams_text_proto, hyperparams_proto)
def mock_conv_argscope_builder(conv_hyperparams_arg, is_training):
return (conv_hyperparams_arg, is_training)
box_predictor_proto = box_predictor_pb2.BoxPredictor()
box_predictor_proto.rfcn_box_predictor.conv_hyperparams.CopyFrom(
hyperparams_proto)
box_predictor = box_predictor_builder.build(
argscope_fn=mock_conv_argscope_builder,
box_predictor_config=box_predictor_proto,
is_training=False,
num_classes=10)
(conv_hyperparams_actual, is_training) = box_predictor._conv_hyperparams
self.assertAlmostEqual((hyperparams_proto.regularizer.
l1_regularizer.weight),
(conv_hyperparams_actual.regularizer.l1_regularizer.
weight))
self.assertAlmostEqual((hyperparams_proto.initializer.
truncated_normal_initializer.stddev),
(conv_hyperparams_actual.initializer.
truncated_normal_initializer.stddev))
self.assertAlmostEqual((hyperparams_proto.initializer.
truncated_normal_initializer.mean),
(conv_hyperparams_actual.initializer.
truncated_normal_initializer.mean))
self.assertEqual(hyperparams_proto.activation,
conv_hyperparams_actual.activation)
self.assertFalse(is_training)
def test_non_default_rfcn_box_predictor(self):
conv_hyperparams_text_proto = """
regularizer {
l1_regularizer {
}
}
initializer {
truncated_normal_initializer {
}
}
activation: RELU_6
"""
box_predictor_text_proto = """
rfcn_box_predictor {
num_spatial_bins_height: 4
num_spatial_bins_width: 4
depth: 4
box_code_size: 3
crop_height: 16
crop_width: 16
}
"""
hyperparams_proto = hyperparams_pb2.Hyperparams()
text_format.Merge(conv_hyperparams_text_proto, hyperparams_proto)
def mock_conv_argscope_builder(conv_hyperparams_arg, is_training):
return (conv_hyperparams_arg, is_training)
box_predictor_proto = box_predictor_pb2.BoxPredictor()
text_format.Merge(box_predictor_text_proto, box_predictor_proto)
box_predictor_proto.rfcn_box_predictor.conv_hyperparams.CopyFrom(
hyperparams_proto)
box_predictor = box_predictor_builder.build(
argscope_fn=mock_conv_argscope_builder,
box_predictor_config=box_predictor_proto,
is_training=True,
num_classes=90)
self.assertEqual(box_predictor.num_classes, 90)
self.assertTrue(box_predictor._is_training)
self.assertEqual(box_predictor._box_code_size, 3)
self.assertEqual(box_predictor._num_spatial_bins, [4, 4])
self.assertEqual(box_predictor._crop_size, [16, 16])
def test_default_rfcn_box_predictor(self):
conv_hyperparams_text_proto = """
regularizer {
l1_regularizer {
}
}
initializer {
truncated_normal_initializer {
}
}
activation: RELU_6
"""
hyperparams_proto = hyperparams_pb2.Hyperparams()
text_format.Merge(conv_hyperparams_text_proto, hyperparams_proto)
def mock_conv_argscope_builder(conv_hyperparams_arg, is_training):
return (conv_hyperparams_arg, is_training)
box_predictor_proto = box_predictor_pb2.BoxPredictor()
box_predictor_proto.rfcn_box_predictor.conv_hyperparams.CopyFrom(
hyperparams_proto)
box_predictor = box_predictor_builder.build(
argscope_fn=mock_conv_argscope_builder,
box_predictor_config=box_predictor_proto,
is_training=True,
num_classes=90)
self.assertEqual(box_predictor.num_classes, 90)
self.assertTrue(box_predictor._is_training)
self.assertEqual(box_predictor._box_code_size, 4)
self.assertEqual(box_predictor._num_spatial_bins, [3, 3])
self.assertEqual(box_predictor._crop_size, [12, 12])
if __name__ == '__main__':
tf.test.main()
object_detection/builders/hyperparams_builder.py 0 → 100644
View file @ a4944a57
# Copyright 2017 The TensorFlow Authors. All Rights Reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
# ==============================================================================
"""Builder function to construct tf-slim arg_scope for convolution, fc ops."""
import tensorflow as tf
from object_detection.protos import hyperparams_pb2
slim = tf.contrib.slim
def build(hyperparams_config, is_training):
"""Builds tf-slim arg_scope for convolution ops based on the config.
Returns an arg_scope to use for convolution ops containing weights
initializer, weights regularizer, activation function, batch norm function
and batch norm parameters based on the configuration.
Note that if the batch_norm parameteres are not specified in the config
(i.e. left to default) then batch norm is excluded from the arg_scope.
The batch norm parameters are set for updates based on `is_training` argument
and conv_hyperparams_config.batch_norm.train parameter. During training, they
are updated only if batch_norm.train parameter is true. However, during eval,
no updates are made to the batch norm variables. In both cases, their current
values are used during forward pass.
Args:
hyperparams_config: hyperparams.proto object containing
hyperparameters.
is_training: Whether the network is in training mode.
Returns:
arg_scope: tf-slim arg_scope containing hyperparameters for ops.
Raises:
ValueError: if hyperparams_config is not of type hyperparams.Hyperparams.
"""
if not isinstance(hyperparams_config,
hyperparams_pb2.Hyperparams):
raise ValueError('hyperparams_config not of type '
'hyperparams_pb.Hyperparams.')
batch_norm = None
batch_norm_params = None
if hyperparams_config.HasField('batch_norm'):
batch_norm = slim.batch_norm
batch_norm_params = _build_batch_norm_params(
hyperparams_config.batch_norm, is_training)
affected_ops = [slim.conv2d, slim.separable_conv2d, slim.conv2d_transpose]
if hyperparams_config.HasField('op') and (
hyperparams_config.op == hyperparams_pb2.Hyperparams.FC):
affected_ops = [slim.fully_connected]
with slim.arg_scope(
affected_ops,
weights_regularizer=_build_regularizer(
hyperparams_config.regularizer),
weights_initializer=_build_initializer(
hyperparams_config.initializer),
activation_fn=_build_activation_fn(hyperparams_config.activation),
normalizer_fn=batch_norm,
normalizer_params=batch_norm_params) as sc:
return sc
def _build_activation_fn(activation_fn):
"""Builds a callable activation from config.
Args:
activation_fn: hyperparams_pb2.Hyperparams.activation
Returns:
Callable activation function.
Raises:
ValueError: On unknown activation function.
"""
if activation_fn == hyperparams_pb2.Hyperparams.NONE:
return None
if activation_fn == hyperparams_pb2.Hyperparams.RELU:
return tf.nn.relu
if activation_fn == hyperparams_pb2.Hyperparams.RELU_6:
return tf.nn.relu6
raise ValueError('Unknown activation function: {}'.format(activation_fn))
def _build_regularizer(regularizer):
"""Builds a tf-slim regularizer from config.
Args:
regularizer: hyperparams_pb2.Hyperparams.regularizer proto.
Returns:
tf-slim regularizer.
Raises:
ValueError: On unknown regularizer.
"""
regularizer_oneof = regularizer.WhichOneof('regularizer_oneof')
if regularizer_oneof == 'l1_regularizer':
return slim.l1_regularizer(scale=regularizer.l1_regularizer.weight)
if regularizer_oneof == 'l2_regularizer':
return slim.l2_regularizer(scale=regularizer.l2_regularizer.weight)
raise ValueError('Unknown regularizer function: {}'.format(regularizer_oneof))
def _build_initializer(initializer):
"""Build a tf initializer from config.
Args:
initializer: hyperparams_pb2.Hyperparams.regularizer proto.
Returns:
tf initializer.
Raises:
ValueError: On unknown initializer.
"""
initializer_oneof = initializer.WhichOneof('initializer_oneof')
if initializer_oneof == 'truncated_normal_initializer':
return tf.truncated_normal_initializer(
mean=initializer.truncated_normal_initializer.mean,
stddev=initializer.truncated_normal_initializer.stddev)
if initializer_oneof == 'variance_scaling_initializer':
enum_descriptor = (hyperparams_pb2.VarianceScalingInitializer.
DESCRIPTOR.enum_types_by_name['Mode'])
mode = enum_descriptor.values_by_number[initializer.
variance_scaling_initializer.
mode].name
return slim.variance_scaling_initializer(
factor=initializer.variance_scaling_initializer.factor,
mode=mode,
uniform=initializer.variance_scaling_initializer.uniform)
raise ValueError('Unknown initializer function: {}'.format(
initializer_oneof))
def _build_batch_norm_params(batch_norm, is_training):
"""Build a dictionary of batch_norm params from config.
Args:
batch_norm: hyperparams_pb2.ConvHyperparams.batch_norm proto.
is_training: Whether the models is in training mode.
Returns:
A dictionary containing batch_norm parameters.
"""
batch_norm_params = {
'decay': batch_norm.decay,
'center': batch_norm.center,
'scale': batch_norm.scale,
'epsilon': batch_norm.epsilon,
'fused': True,
'is_training': is_training and batch_norm.train,
}
return batch_norm_params
object_detection/builders/hyperparams_builder_test.py 0 → 100644
View file @ a4944a57
# Copyright 2017 The TensorFlow Authors. All Rights Reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
# ==============================================================================
"""Tests object_detection.core.hyperparams_builder."""
import numpy as np
import tensorflow as tf
from google.protobuf import text_format
# TODO: Rewrite third_party imports.
from object_detection.builders import hyperparams_builder
from object_detection.protos import hyperparams_pb2
slim = tf.contrib.slim
class HyperparamsBuilderTest(tf.test.TestCase):
# TODO: Make this a public api in slim arg_scope.py.
def _get_scope_key(self, op):
return getattr(op, '_key_op', str(op))
def test_default_arg_scope_has_conv2d_op(self):
conv_hyperparams_text_proto = """
regularizer {
l1_regularizer {
}
}
initializer {
truncated_normal_initializer {
}
}
"""
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)
self.assertTrue(self._get_scope_key(slim.conv2d) in scope)
def test_default_arg_scope_has_separable_conv2d_op(self):
conv_hyperparams_text_proto = """
regularizer {
l1_regularizer {
}
}
initializer {
truncated_normal_initializer {
}
}
"""
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)
self.assertTrue(self._get_scope_key(slim.separable_conv2d) in scope)
def test_default_arg_scope_has_conv2d_transpose_op(self):
conv_hyperparams_text_proto = """
regularizer {
l1_regularizer {
}
}
initializer {
truncated_normal_initializer {
}
}
"""
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)
self.assertTrue(self._get_scope_key(slim.conv2d_transpose) in scope)
def test_explicit_fc_op_arg_scope_has_fully_connected_op(self):
conv_hyperparams_text_proto = """
op: FC
regularizer {
l1_regularizer {
}
}
initializer {
truncated_normal_initializer {
}
}
"""
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)
self.assertTrue(self._get_scope_key(slim.fully_connected) in scope)
def test_separable_conv2d_and_conv2d_and_transpose_have_same_parameters(self):
conv_hyperparams_text_proto = """
regularizer {
l1_regularizer {
}
}
initializer {
truncated_normal_initializer {
}
}
"""
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)
kwargs_1, kwargs_2, kwargs_3 = scope.values()
self.assertDictEqual(kwargs_1, kwargs_2)
self.assertDictEqual(kwargs_1, kwargs_3)
def test_return_l1_regularized_weights(self):
conv_hyperparams_text_proto = """
regularizer {
l1_regularizer {
weight: 0.5
}
}
initializer {
truncated_normal_initializer {
}
}
"""
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]
regularizer = conv_scope_arguments['weights_regularizer']
weights = np.array([1., -1, 4., 2.])
with self.test_session() as sess:
result = sess.run(regularizer(tf.constant(weights)))
self.assertAllClose(np.abs(weights).sum() * 0.5, result)
def test_return_l2_regularizer_weights(self):
conv_hyperparams_text_proto = """
regularizer {
l2_regularizer {
weight: 0.42
}
}
initializer {
truncated_normal_initializer {
}
}
"""
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]
regularizer = conv_scope_arguments['weights_regularizer']
weights = np.array([1., -1, 4., 2.])
with self.test_session() as sess:
result = sess.run(regularizer(tf.constant(weights)))
self.assertAllClose(np.power(weights, 2).sum() / 2.0 * 0.42, result)
def test_return_non_default_batch_norm_params_with_train_during_train(self):
conv_hyperparams_text_proto = """
regularizer {
l2_regularizer {
}
}
initializer {
truncated_normal_initializer {
}
}
batch_norm {
decay: 0.7
center: false
scale: true
epsilon: 0.03
train: true
}
"""
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]
self.assertEqual(conv_scope_arguments['normalizer_fn'], slim.batch_norm)
batch_norm_params = conv_scope_arguments['normalizer_params']
self.assertAlmostEqual(batch_norm_params['decay'], 0.7)
self.assertAlmostEqual(batch_norm_params['epsilon'], 0.03)
self.assertFalse(batch_norm_params['center'])
self.assertTrue(batch_norm_params['scale'])
self.assertTrue(batch_norm_params['is_training'])
def test_return_batch_norm_params_with_notrain_during_eval(self):
conv_hyperparams_text_proto = """
regularizer {
l2_regularizer {
}
}
initializer {
truncated_normal_initializer {
}
}
batch_norm {
decay: 0.7
center: false
scale: true
epsilon: 0.03
train: true
}
"""
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=False)
conv_scope_arguments = scope.values()[0]
self.assertEqual(conv_scope_arguments['normalizer_fn'], slim.batch_norm)
batch_norm_params = conv_scope_arguments['normalizer_params']
self.assertAlmostEqual(batch_norm_params['decay'], 0.7)
self.assertAlmostEqual(batch_norm_params['epsilon'], 0.03)
self.assertFalse(batch_norm_params['center'])
self.assertTrue(batch_norm_params['scale'])
self.assertFalse(batch_norm_params['is_training'])
def test_return_batch_norm_params_with_notrain_when_train_is_false(self):
conv_hyperparams_text_proto = """
regularizer {
l2_regularizer {
}
}
initializer {
truncated_normal_initializer {
}
}
batch_norm {
decay: 0.7
center: false
scale: true
epsilon: 0.03
train: false
}
"""
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]
self.assertEqual(conv_scope_arguments['normalizer_fn'], slim.batch_norm)
batch_norm_params = conv_scope_arguments['normalizer_params']
self.assertAlmostEqual(batch_norm_params['decay'], 0.7)
self.assertAlmostEqual(batch_norm_params['epsilon'], 0.03)
self.assertFalse(batch_norm_params['center'])
self.assertTrue(batch_norm_params['scale'])
self.assertFalse(batch_norm_params['is_training'])
def test_do_not_use_batch_norm_if_default(self):
conv_hyperparams_text_proto = """
regularizer {
l2_regularizer {
}
}
initializer {
truncated_normal_initializer {
}
}
"""
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]
self.assertEqual(conv_scope_arguments['normalizer_fn'], None)
self.assertEqual(conv_scope_arguments['normalizer_params'], None)
def test_use_none_activation(self):
conv_hyperparams_text_proto = """
regularizer {
l2_regularizer {
}
}
initializer {
truncated_normal_initializer {
}
}
activation: NONE
"""
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]
self.assertEqual(conv_scope_arguments['activation_fn'], None)
def test_use_relu_activation(self):
conv_hyperparams_text_proto = """
regularizer {
l2_regularizer {
}
}
initializer {
truncated_normal_initializer {
}
}
activation: RELU
"""
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]
self.assertEqual(conv_scope_arguments['activation_fn'], tf.nn.relu)
def test_use_relu_6_activation(self):
conv_hyperparams_text_proto = """
regularizer {
l2_regularizer {
}
}
initializer {
truncated_normal_initializer {
}
}
activation: RELU_6
"""
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]
self.assertEqual(conv_scope_arguments['activation_fn'], tf.nn.relu6)
def _assert_variance_in_range(self, initializer, shape, variance,
tol=1e-2):
with tf.Graph().as_default() as g:
with self.test_session(graph=g) as sess:
var = tf.get_variable(
name='test',
shape=shape,
dtype=tf.float32,
initializer=initializer)
sess.run(tf.global_variables_initializer())
values = sess.run(var)
self.assertAllClose(np.var(values), variance, tol, tol)
def test_variance_in_range_with_variance_scaling_initializer_fan_in(self):
conv_hyperparams_text_proto = """
regularizer {
l2_regularizer {
}
}
initializer {
variance_scaling_initializer {
factor: 2.0
mode: FAN_IN
uniform: false
}
}
"""
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=2. / 100.)
def test_variance_in_range_with_variance_scaling_initializer_fan_out(self):
conv_hyperparams_text_proto = """
regularizer {
l2_regularizer {
}
}
initializer {
variance_scaling_initializer {
factor: 2.0
mode: FAN_OUT
uniform: false
}
}
"""
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=2. / 40.)
def test_variance_in_range_with_variance_scaling_initializer_fan_avg(self):
conv_hyperparams_text_proto = """
regularizer {
l2_regularizer {
}
}
initializer {
variance_scaling_initializer {
factor: 2.0
mode: FAN_AVG
uniform: false
}
}
"""
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=4. / (100. + 40.))
def test_variance_in_range_with_variance_scaling_initializer_uniform(self):
conv_hyperparams_text_proto = """
regularizer {
l2_regularizer {
}
}
initializer {
variance_scaling_initializer {
factor: 2.0
mode: FAN_IN
uniform: true
}
}
"""
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=2. / 100.)
def test_variance_in_range_with_truncated_normal_initializer(self):
conv_hyperparams_text_proto = """
regularizer {
l2_regularizer {
}
}
initializer {
truncated_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=0.49, tol=1e-1)
if __name__ == '__main__':
tf.test.main()
object_detection/builders/image_resizer_builder.py 0 → 100644
View file @ a4944a57
# Copyright 2017 The TensorFlow Authors. All Rights Reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
# ==============================================================================
"""Builder function for image resizing operations."""
import functools
from object_detection.core import preprocessor
from object_detection.protos import image_resizer_pb2
def build(image_resizer_config):
"""Builds callable for image resizing operations.
Args:
image_resizer_config: image_resizer.proto object containing parameters for
an image resizing operation.
Returns:
image_resizer_fn: Callable for image resizing. This callable always takes
a rank-3 image tensor (corresponding to a single image) and returns a
rank-3 image tensor, possibly with new spatial dimensions.
Raises:
ValueError: if `image_resizer_config` is of incorrect type.
ValueError: if `image_resizer_config.image_resizer_oneof` is of expected
type.
ValueError: if min_dimension > max_dimension when keep_aspect_ratio_resizer
is used.
"""
if not isinstance(image_resizer_config, image_resizer_pb2.ImageResizer):
raise ValueError('image_resizer_config not of type '
'image_resizer_pb2.ImageResizer.')
if image_resizer_config.WhichOneof(
'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):
raise ValueError('min_dimension > max_dimension')
return functools.partial(
preprocessor.resize_to_range,
min_dimension=keep_aspect_ratio_config.min_dimension,
max_dimension=keep_aspect_ratio_config.max_dimension)
if image_resizer_config.WhichOneof(
'image_resizer_oneof') == 'fixed_shape_resizer':
fixed_shape_resizer_config = image_resizer_config.fixed_shape_resizer
return functools.partial(preprocessor.resize_image,
new_height=fixed_shape_resizer_config.height,
new_width=fixed_shape_resizer_config.width)
raise ValueError('Invalid image resizer option.')
object_detection/builders/image_resizer_builder_test.py 0 → 100644
View file @ a4944a57
# Copyright 2017 The TensorFlow Authors. All Rights Reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
# ==============================================================================
"""Tests for object_detection.builders.image_resizer_builder."""
import tensorflow as tf
from google.protobuf import text_format
from object_detection.builders import image_resizer_builder
from object_detection.protos import image_resizer_pb2
class ImageResizerBuilderTest(tf.test.TestCase):
def _shape_of_resized_random_image_given_text_proto(
self, input_shape, text_proto):
image_resizer_config = image_resizer_pb2.ImageResizer()
text_format.Merge(text_proto, image_resizer_config)
image_resizer_fn = image_resizer_builder.build(image_resizer_config)
images = tf.to_float(tf.random_uniform(
input_shape, minval=0, maxval=255, dtype=tf.int32))
resized_images = image_resizer_fn(images)
with self.test_session() as sess:
return sess.run(resized_images).shape
def test_built_keep_aspect_ratio_resizer_returns_expected_shape(self):
image_resizer_text_proto = """
keep_aspect_ratio_resizer {
min_dimension: 10
max_dimension: 20
}
"""
input_shape = (50, 25, 3)
expected_output_shape = (20, 10, 3)
output_shape = self._shape_of_resized_random_image_given_text_proto(
input_shape, image_resizer_text_proto)
self.assertEqual(output_shape, expected_output_shape)
def test_built_fixed_shape_resizer_returns_expected_shape(self):
image_resizer_text_proto = """
fixed_shape_resizer {
height: 10
width: 20
}
"""
input_shape = (50, 25, 3)
expected_output_shape = (10, 20, 3)
output_shape = self._shape_of_resized_random_image_given_text_proto(
input_shape, image_resizer_text_proto)
self.assertEqual(output_shape, expected_output_shape)
def test_raises_error_on_invalid_input(self):
invalid_input = 'invalid_input'
with self.assertRaises(ValueError):
image_resizer_builder.build(invalid_input)
if __name__ == '__main__':
tf.test.main()
object_detection/builders/input_reader_builder.py 0 → 100644
View file @ a4944a57
# Copyright 2017 The TensorFlow Authors. All Rights Reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
# ==============================================================================
"""Input reader builder.
Creates data sources for DetectionModels from an InputReader config. See
input_reader.proto for options.
Note: If users wishes to also use their own InputReaders with the Object
Detection configuration framework, they should define their own builder function
that wraps the build function.
"""
import tensorflow as tf
from object_detection.data_decoders import tf_example_decoder
from object_detection.protos import input_reader_pb2
parallel_reader = tf.contrib.slim.parallel_reader
def build(input_reader_config):
"""Builds a tensor dictionary based on the InputReader config.
Args:
input_reader_config: A input_reader_pb2.InputReader object.
Returns:
A tensor dict based on the input_reader_config.
Raises:
ValueError: On invalid input reader proto.
"""
if not isinstance(input_reader_config, input_reader_pb2.InputReader):
raise ValueError('input_reader_config not of type '
'input_reader_pb2.InputReader.')
if input_reader_config.WhichOneof('input_reader') == 'tf_record_input_reader':
config = input_reader_config.tf_record_input_reader
_, string_tensor = parallel_reader.parallel_read(
config.input_path,
reader_class=tf.TFRecordReader,
num_epochs=(input_reader_config.num_epochs
if input_reader_config.num_epochs else None),
num_readers=input_reader_config.num_readers,
shuffle=input_reader_config.shuffle,
dtypes=[tf.string, tf.string],
capacity=input_reader_config.queue_capacity,
min_after_dequeue=input_reader_config.min_after_dequeue)
return tf_example_decoder.TfExampleDecoder().Decode(string_tensor)
raise ValueError('Unsupported input_reader_config.')
object_detection/builders/input_reader_builder_test.py 0 → 100644
View file @ a4944a57
# Copyright 2017 The TensorFlow Authors. All Rights Reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
# ==============================================================================
"""Tests for input_reader_builder."""
import os
import numpy as np
import tensorflow as tf
from google.protobuf import text_format
from tensorflow.core.example import example_pb2
from tensorflow.core.example import feature_pb2
from object_detection.builders import input_reader_builder
from object_detection.core import standard_fields as fields
from object_detection.protos import input_reader_pb2
class InputReaderBuilderTest(tf.test.TestCase):
def create_tf_record(self):
path = os.path.join(self.get_temp_dir(), 'tfrecord')
writer = tf.python_io.TFRecordWriter(path)
image_tensor = np.random.randint(255, size=(4, 5, 3)).astype(np.uint8)
with self.test_session():
encoded_jpeg = tf.image.encode_jpeg(tf.constant(image_tensor)).eval()
example = example_pb2.Example(features=feature_pb2.Features(feature={
'image/encoded': feature_pb2.Feature(
bytes_list=feature_pb2.BytesList(value=[encoded_jpeg])),
'image/format': feature_pb2.Feature(
bytes_list=feature_pb2.BytesList(value=['jpeg'.encode('utf-8')])),
'image/object/bbox/xmin': feature_pb2.Feature(
float_list=feature_pb2.FloatList(value=[0.0])),
'image/object/bbox/xmax': feature_pb2.Feature(
float_list=feature_pb2.FloatList(value=[1.0])),
'image/object/bbox/ymin': feature_pb2.Feature(
float_list=feature_pb2.FloatList(value=[0.0])),
'image/object/bbox/ymax': feature_pb2.Feature(
float_list=feature_pb2.FloatList(value=[1.0])),
'image/object/class/label': feature_pb2.Feature(
int64_list=feature_pb2.Int64List(value=[2])),
}))
writer.write(example.SerializeToString())
writer.close()
return path
def test_build_tf_record_input_reader(self):
tf_record_path = self.create_tf_record()
input_reader_text_proto = """
shuffle: false
num_readers: 1
tf_record_input_reader {{
input_path: '{0}'
}}
""".format(tf_record_path)
input_reader_proto = input_reader_pb2.InputReader()
text_format.Merge(input_reader_text_proto, input_reader_proto)
tensor_dict = input_reader_builder.build(input_reader_proto)
sv = tf.train.Supervisor(logdir=self.get_temp_dir())
with sv.prepare_or_wait_for_session() as sess:
sv.start_queue_runners(sess)
output_dict = sess.run(tensor_dict)
self.assertEquals(
(4, 5, 3), output_dict[fields.InputDataFields.image].shape)
self.assertEquals(
[2], output_dict[fields.InputDataFields.groundtruth_classes])
self.assertEquals(
(1, 4), output_dict[fields.InputDataFields.groundtruth_boxes].shape)
self.assertAllEqual(
[0.0, 0.0, 1.0, 1.0],
output_dict[fields.InputDataFields.groundtruth_boxes][0])
if __name__ == '__main__':
tf.test.main()
object_detection/builders/losses_builder.py 0 → 100644
View file @ a4944a57
# Copyright 2017 The TensorFlow Authors. All Rights Reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
# ==============================================================================
"""A function to build localization and classification losses from config."""
from object_detection.core import losses
from object_detection.protos import losses_pb2
def build(loss_config):
"""Build losses based on the config.
Builds classification, localization losses and optionally a hard example miner
based on the config.
Args:
loss_config: A losses_pb2.Loss object.
Returns:
classification_loss: Classification loss object.
localization_loss: Localization loss object.
classification_weight: Classification loss weight.
localization_weight: Localization loss weight.
hard_example_miner: Hard example miner object.
"""
classification_loss = _build_classification_loss(
loss_config.classification_loss)
localization_loss = _build_localization_loss(
loss_config.localization_loss)
classification_weight = loss_config.classification_weight
localization_weight = loss_config.localization_weight
hard_example_miner = None
if loss_config.HasField('hard_example_miner'):
hard_example_miner = build_hard_example_miner(
loss_config.hard_example_miner,
classification_weight,
localization_weight)
return (classification_loss, localization_loss,
classification_weight,
localization_weight, hard_example_miner)
def build_hard_example_miner(config,
classification_weight,
localization_weight):
"""Builds hard example miner based on the config.
Args:
config: A losses_pb2.HardExampleMiner object.
classification_weight: Classification loss weight.
localization_weight: Localization loss weight.
Returns:
Hard example miner.
"""
loss_type = None
if config.loss_type == losses_pb2.HardExampleMiner.BOTH:
loss_type = 'both'
if config.loss_type == losses_pb2.HardExampleMiner.CLASSIFICATION:
loss_type = 'cls'
if config.loss_type == losses_pb2.HardExampleMiner.LOCALIZATION:
loss_type = 'loc'
max_negatives_per_positive = None
num_hard_examples = None
if config.max_negatives_per_positive > 0:
max_negatives_per_positive = config.max_negatives_per_positive
if config.num_hard_examples > 0:
num_hard_examples = config.num_hard_examples
hard_example_miner = losses.HardExampleMiner(
num_hard_examples=num_hard_examples,
iou_threshold=config.iou_threshold,
loss_type=loss_type,
cls_loss_weight=classification_weight,
loc_loss_weight=localization_weight,
max_negatives_per_positive=max_negatives_per_positive,
min_negatives_per_image=config.min_negatives_per_image)
return hard_example_miner
def _build_localization_loss(loss_config):
"""Builds a localization loss based on the loss config.
Args:
loss_config: A losses_pb2.LocalizationLoss object.
Returns:
Loss based on the config.
Raises:
ValueError: On invalid loss_config.
"""
if not isinstance(loss_config, losses_pb2.LocalizationLoss):
raise ValueError('loss_config not of type losses_pb2.LocalizationLoss.')
loss_type = loss_config.WhichOneof('localization_loss')
if loss_type == 'weighted_l2':
config = loss_config.weighted_l2
return losses.WeightedL2LocalizationLoss(
anchorwise_output=config.anchorwise_output)
if loss_type == 'weighted_smooth_l1':
config = loss_config.weighted_smooth_l1
return losses.WeightedSmoothL1LocalizationLoss(
anchorwise_output=config.anchorwise_output)
if loss_type == 'weighted_iou':
return losses.WeightedIOULocalizationLoss()
raise ValueError('Empty loss config.')
def _build_classification_loss(loss_config):
"""Builds a classification loss based on the loss config.
Args:
loss_config: A losses_pb2.ClassificationLoss object.
Returns:
Loss based on the config.
Raises:
ValueError: On invalid loss_config.
"""
if not isinstance(loss_config, losses_pb2.ClassificationLoss):
raise ValueError('loss_config not of type losses_pb2.ClassificationLoss.')
loss_type = loss_config.WhichOneof('classification_loss')
if loss_type == 'weighted_sigmoid':
config = loss_config.weighted_sigmoid
return losses.WeightedSigmoidClassificationLoss(
anchorwise_output=config.anchorwise_output)
if loss_type == 'weighted_softmax':
config = loss_config.weighted_softmax
return losses.WeightedSoftmaxClassificationLoss(
anchorwise_output=config.anchorwise_output)
if loss_type == 'bootstrapped_sigmoid':
config = loss_config.bootstrapped_sigmoid
return losses.BootstrappedSigmoidClassificationLoss(
alpha=config.alpha,
bootstrap_type=('hard' if config.hard_bootstrap else 'soft'),
anchorwise_output=config.anchorwise_output)
raise ValueError('Empty loss config.')
object_detection/builders/losses_builder_test.py 0 → 100644
View file @ a4944a57
# Copyright 2017 The TensorFlow Authors. All Rights Reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
# ==============================================================================
"""Tests for losses_builder."""
import tensorflow as tf
from google.protobuf import text_format
from object_detection.builders import losses_builder
from object_detection.core import losses
from object_detection.protos import losses_pb2
class LocalizationLossBuilderTest(tf.test.TestCase):
def test_build_weighted_l2_localization_loss(self):
losses_text_proto = """
localization_loss {
weighted_l2 {
}
}
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.WeightedL2LocalizationLoss))
def test_build_weighted_smooth_l1_localization_loss(self):
losses_text_proto = """
localization_loss {
weighted_smooth_l1 {
}
}
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))
def test_build_weighted_iou_localization_loss(self):
losses_text_proto = """
localization_loss {
weighted_iou {
}
}
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.WeightedIOULocalizationLoss))
def test_anchorwise_output(self):
losses_text_proto = """
localization_loss {
weighted_smooth_l1 {
anchorwise_output: true
}
}
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))
predictions = tf.constant([[[0.0, 0.0, 1.0, 1.0], [0.0, 0.0, 1.0, 1.0]]])
targets = tf.constant([[[0.0, 0.0, 1.0, 1.0], [0.0, 0.0, 1.0, 1.0]]])
weights = tf.constant([[1.0, 1.0]])
loss = localization_loss(predictions, targets, weights=weights)
self.assertEqual(loss.shape, [1, 2])
def test_raise_error_on_empty_localization_config(self):
losses_text_proto = """
classification_loss {
weighted_softmax {
}
}
"""
losses_proto = losses_pb2.Loss()
text_format.Merge(losses_text_proto, losses_proto)
with self.assertRaises(ValueError):
losses_builder._build_localization_loss(losses_proto)
class ClassificationLossBuilderTest(tf.test.TestCase):
def test_build_weighted_sigmoid_classification_loss(self):
losses_text_proto = """
classification_loss {
weighted_sigmoid {
}
}
localization_loss {
weighted_l2 {
}
}
"""
losses_proto = losses_pb2.Loss()
text_format.Merge(losses_text_proto, losses_proto)
classification_loss, _, _, _, _ = losses_builder.build(losses_proto)
self.assertTrue(isinstance(classification_loss,
losses.WeightedSigmoidClassificationLoss))
def test_build_weighted_softmax_classification_loss(self):
losses_text_proto = """
classification_loss {
weighted_softmax {
}
}
localization_loss {
weighted_l2 {
}
}
"""
losses_proto = losses_pb2.Loss()
text_format.Merge(losses_text_proto, losses_proto)
classification_loss, _, _, _, _ = losses_builder.build(losses_proto)
self.assertTrue(isinstance(classification_loss,
losses.WeightedSoftmaxClassificationLoss))
def test_build_bootstrapped_sigmoid_classification_loss(self):
losses_text_proto = """
classification_loss {
bootstrapped_sigmoid {
alpha: 0.5
}
}
localization_loss {
weighted_l2 {
}
}
"""
losses_proto = losses_pb2.Loss()
text_format.Merge(losses_text_proto, losses_proto)
classification_loss, _, _, _, _ = losses_builder.build(losses_proto)
self.assertTrue(isinstance(classification_loss,
losses.BootstrappedSigmoidClassificationLoss))
def test_anchorwise_output(self):
losses_text_proto = """
classification_loss {
weighted_sigmoid {
anchorwise_output: true
}
}
localization_loss {
weighted_l2 {
}
}
"""
losses_proto = losses_pb2.Loss()
text_format.Merge(losses_text_proto, losses_proto)
classification_loss, _, _, _, _ = losses_builder.build(losses_proto)
self.assertTrue(isinstance(classification_loss,
losses.WeightedSigmoidClassificationLoss))
predictions = tf.constant([[[0.0, 1.0, 0.0], [0.0, 0.5, 0.5]]])
targets = tf.constant([[[0.0, 1.0, 0.0], [0.0, 0.0, 1.0]]])
weights = tf.constant([[1.0, 1.0]])
loss = classification_loss(predictions, targets, weights=weights)
self.assertEqual(loss.shape, [1, 2])
def test_raise_error_on_empty_config(self):
losses_text_proto = """
localization_loss {
weighted_l2 {
}
}
"""
losses_proto = losses_pb2.Loss()
text_format.Merge(losses_text_proto, losses_proto)
with self.assertRaises(ValueError):
losses_builder.build(losses_proto)
class HardExampleMinerBuilderTest(tf.test.TestCase):
def test_do_not_build_hard_example_miner_by_default(self):
losses_text_proto = """
localization_loss {
weighted_l2 {
}
}
classification_loss {
weighted_softmax {
}
}
"""
losses_proto = losses_pb2.Loss()
text_format.Merge(losses_text_proto, losses_proto)
_, _, _, _, hard_example_miner = losses_builder.build(losses_proto)
self.assertEqual(hard_example_miner, None)
def test_build_hard_example_miner_for_classification_loss(self):
losses_text_proto = """
localization_loss {
weighted_l2 {
}
}
classification_loss {
weighted_softmax {
}
}
hard_example_miner {
loss_type: CLASSIFICATION
}
"""
losses_proto = losses_pb2.Loss()
text_format.Merge(losses_text_proto, losses_proto)
_, _, _, _, hard_example_miner = losses_builder.build(losses_proto)
self.assertTrue(isinstance(hard_example_miner, losses.HardExampleMiner))
self.assertEqual(hard_example_miner._loss_type, 'cls')
def test_build_hard_example_miner_for_localization_loss(self):
losses_text_proto = """
localization_loss {
weighted_l2 {
}
}
classification_loss {
weighted_softmax {
}
}
hard_example_miner {
loss_type: LOCALIZATION
}
"""
losses_proto = losses_pb2.Loss()
text_format.Merge(losses_text_proto, losses_proto)
_, _, _, _, hard_example_miner = losses_builder.build(losses_proto)
self.assertTrue(isinstance(hard_example_miner, losses.HardExampleMiner))
self.assertEqual(hard_example_miner._loss_type, 'loc')
def test_build_hard_example_miner_with_non_default_values(self):
losses_text_proto = """
localization_loss {
weighted_l2 {
}
}
classification_loss {
weighted_softmax {
}
}
hard_example_miner {
num_hard_examples: 32
iou_threshold: 0.5
loss_type: LOCALIZATION
max_negatives_per_positive: 10
min_negatives_per_image: 3
}
"""
losses_proto = losses_pb2.Loss()
text_format.Merge(losses_text_proto, losses_proto)
_, _, _, _, hard_example_miner = losses_builder.build(losses_proto)
self.assertTrue(isinstance(hard_example_miner, losses.HardExampleMiner))
self.assertEqual(hard_example_miner._num_hard_examples, 32)
self.assertAlmostEqual(hard_example_miner._iou_threshold, 0.5)
self.assertEqual(hard_example_miner._max_negatives_per_positive, 10)
self.assertEqual(hard_example_miner._min_negatives_per_image, 3)
class LossBuilderTest(tf.test.TestCase):
def test_build_all_loss_parameters(self):
losses_text_proto = """
localization_loss {
weighted_l2 {
}
}
classification_loss {
weighted_softmax {
}
}
hard_example_miner {
}
classification_weight: 0.8
localization_weight: 0.2
"""
losses_proto = losses_pb2.Loss()
text_format.Merge(losses_text_proto, losses_proto)
(classification_loss, localization_loss,
classification_weight, localization_weight,
hard_example_miner) = losses_builder.build(losses_proto)
self.assertTrue(isinstance(hard_example_miner, losses.HardExampleMiner))
self.assertTrue(isinstance(classification_loss,
losses.WeightedSoftmaxClassificationLoss))
self.assertTrue(isinstance(localization_loss,
losses.WeightedL2LocalizationLoss))
self.assertAlmostEqual(classification_weight, 0.8)
self.assertAlmostEqual(localization_weight, 0.2)
if __name__ == '__main__':
tf.test.main()
object_detection/builders/matcher_builder.py 0 → 100644
View file @ a4944a57
# Copyright 2017 The TensorFlow Authors. All Rights Reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
# ==============================================================================
"""A function to build an object detection matcher from configuration."""
from object_detection.matchers import argmax_matcher
from object_detection.matchers import bipartite_matcher
from object_detection.protos import matcher_pb2
def build(matcher_config):
"""Builds a matcher object based on the matcher config.
Args:
matcher_config: A matcher.proto object containing the config for the desired
Matcher.
Returns:
Matcher based on the config.
Raises:
ValueError: On empty matcher proto.
"""
if not isinstance(matcher_config, matcher_pb2.Matcher):
raise ValueError('matcher_config not of type matcher_pb2.Matcher.')
if matcher_config.WhichOneof('matcher_oneof') == 'argmax_matcher':
matcher = matcher_config.argmax_matcher
matched_threshold = unmatched_threshold = None
if not matcher.ignore_thresholds:
matched_threshold = matcher.matched_threshold
unmatched_threshold = matcher.unmatched_threshold
return argmax_matcher.ArgMaxMatcher(
matched_threshold=matched_threshold,
unmatched_threshold=unmatched_threshold,
negatives_lower_than_unmatched=matcher.negatives_lower_than_unmatched,
force_match_for_each_row=matcher.force_match_for_each_row)
if matcher_config.WhichOneof('matcher_oneof') == 'bipartite_matcher':
return bipartite_matcher.GreedyBipartiteMatcher()
raise ValueError('Empty matcher.')
object_detection/builders/matcher_builder_test.py 0 → 100644
View file @ a4944a57
# Copyright 2017 The TensorFlow Authors. All Rights Reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
# ==============================================================================
"""Tests for matcher_builder."""
import tensorflow as tf
from google.protobuf import text_format
from object_detection.builders import matcher_builder
from object_detection.matchers import argmax_matcher
from object_detection.matchers import bipartite_matcher
from object_detection.protos import matcher_pb2
class MatcherBuilderTest(tf.test.TestCase):
def test_build_arg_max_matcher_with_defaults(self):
matcher_text_proto = """
argmax_matcher {
}
"""
matcher_proto = matcher_pb2.Matcher()
text_format.Merge(matcher_text_proto, matcher_proto)
matcher_object = matcher_builder.build(matcher_proto)
self.assertTrue(isinstance(matcher_object, argmax_matcher.ArgMaxMatcher))
self.assertAlmostEqual(matcher_object._matched_threshold, 0.5)
self.assertAlmostEqual(matcher_object._unmatched_threshold, 0.5)
self.assertTrue(matcher_object._negatives_lower_than_unmatched)
self.assertFalse(matcher_object._force_match_for_each_row)
def test_build_arg_max_matcher_without_thresholds(self):
matcher_text_proto = """
argmax_matcher {
ignore_thresholds: true
}
"""
matcher_proto = matcher_pb2.Matcher()
text_format.Merge(matcher_text_proto, matcher_proto)
matcher_object = matcher_builder.build(matcher_proto)
self.assertTrue(isinstance(matcher_object, argmax_matcher.ArgMaxMatcher))
self.assertEqual(matcher_object._matched_threshold, None)
self.assertEqual(matcher_object._unmatched_threshold, None)
self.assertTrue(matcher_object._negatives_lower_than_unmatched)
self.assertFalse(matcher_object._force_match_for_each_row)
def test_build_arg_max_matcher_with_non_default_parameters(self):
matcher_text_proto = """
argmax_matcher {
matched_threshold: 0.7
unmatched_threshold: 0.3
negatives_lower_than_unmatched: false
force_match_for_each_row: true
}
"""
matcher_proto = matcher_pb2.Matcher()
text_format.Merge(matcher_text_proto, matcher_proto)
matcher_object = matcher_builder.build(matcher_proto)
self.assertTrue(isinstance(matcher_object, argmax_matcher.ArgMaxMatcher))
self.assertAlmostEqual(matcher_object._matched_threshold, 0.7)
self.assertAlmostEqual(matcher_object._unmatched_threshold, 0.3)
self.assertFalse(matcher_object._negatives_lower_than_unmatched)
self.assertTrue(matcher_object._force_match_for_each_row)
def test_build_bipartite_matcher(self):
matcher_text_proto = """
bipartite_matcher {
}
"""
matcher_proto = matcher_pb2.Matcher()
text_format.Merge(matcher_text_proto, matcher_proto)
matcher_object = matcher_builder.build(matcher_proto)
self.assertTrue(
isinstance(matcher_object, bipartite_matcher.GreedyBipartiteMatcher))
def test_raise_error_on_empty_matcher(self):
matcher_text_proto = """
"""
matcher_proto = matcher_pb2.Matcher()
text_format.Merge(matcher_text_proto, matcher_proto)
with self.assertRaises(ValueError):
matcher_builder.build(matcher_proto)
if __name__ == '__main__':
tf.test.main()
object_detection/builders/model_builder.py 0 → 100644
View file @ a4944a57
# Copyright 2017 The TensorFlow Authors. All Rights Reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
# ==============================================================================
"""A function to build a DetectionModel from configuration."""
from object_detection.builders import anchor_generator_builder
from object_detection.builders import box_coder_builder
from object_detection.builders import box_predictor_builder
from object_detection.builders import hyperparams_builder
from object_detection.builders import image_resizer_builder
from object_detection.builders import losses_builder
from object_detection.builders import matcher_builder
from object_detection.builders import post_processing_builder
from object_detection.builders import region_similarity_calculator_builder as sim_calc
from object_detection.core import box_predictor
from object_detection.meta_architectures import faster_rcnn_meta_arch
from object_detection.meta_architectures import rfcn_meta_arch
from object_detection.meta_architectures import ssd_meta_arch
from object_detection.models import faster_rcnn_inception_resnet_v2_feature_extractor as frcnn_inc_res
from object_detection.models import faster_rcnn_resnet_v1_feature_extractor as frcnn_resnet_v1
from object_detection.models.ssd_inception_v2_feature_extractor import SSDInceptionV2FeatureExtractor
from object_detection.models.ssd_mobilenet_v1_feature_extractor import SSDMobileNetV1FeatureExtractor
from object_detection.protos import model_pb2
# A map of names to SSD feature extractors.
SSD_FEATURE_EXTRACTOR_CLASS_MAP = {
'ssd_inception_v2': SSDInceptionV2FeatureExtractor,
'ssd_mobilenet_v1': SSDMobileNetV1FeatureExtractor,
}
# A map of names to Faster R-CNN feature extractors.
FASTER_RCNN_FEATURE_EXTRACTOR_CLASS_MAP = {
'faster_rcnn_resnet50':
frcnn_resnet_v1.FasterRCNNResnet50FeatureExtractor,
'faster_rcnn_resnet101':
frcnn_resnet_v1.FasterRCNNResnet101FeatureExtractor,
'faster_rcnn_resnet152':
frcnn_resnet_v1.FasterRCNNResnet152FeatureExtractor,
'faster_rcnn_inception_resnet_v2':
frcnn_inc_res.FasterRCNNInceptionResnetV2FeatureExtractor
}
def build(model_config, is_training):
"""Builds a DetectionModel based on the model config.
Args:
model_config: A model.proto object containing the config for the desired
DetectionModel.
is_training: True if this model is being built for training purposes.
Returns:
DetectionModel based on the config.
Raises:
ValueError: On invalid meta architecture or model.
"""
if not isinstance(model_config, model_pb2.DetectionModel):
raise ValueError('model_config not of type model_pb2.DetectionModel.')
meta_architecture = model_config.WhichOneof('model')
if meta_architecture == 'ssd':
return _build_ssd_model(model_config.ssd, is_training)
if meta_architecture == 'faster_rcnn':
return _build_faster_rcnn_model(model_config.faster_rcnn, is_training)
raise ValueError('Unknown meta architecture: {}'.format(meta_architecture))
def _build_ssd_feature_extractor(feature_extractor_config, is_training,
reuse_weights=None):
"""Builds a ssd_meta_arch.SSDFeatureExtractor based on config.
Args:
feature_extractor_config: A SSDFeatureExtractor proto config from ssd.proto.
is_training: True if this feature extractor is being built for training.
reuse_weights: if the feature extractor should reuse weights.
Returns:
ssd_meta_arch.SSDFeatureExtractor based on config.
Raises:
ValueError: On invalid feature extractor type.
"""
feature_type = feature_extractor_config.type
depth_multiplier = feature_extractor_config.depth_multiplier
min_depth = feature_extractor_config.min_depth
conv_hyperparams = hyperparams_builder.build(
feature_extractor_config.conv_hyperparams, is_training)
if feature_type not in SSD_FEATURE_EXTRACTOR_CLASS_MAP:
raise ValueError('Unknown ssd feature_extractor: {}'.format(feature_type))
feature_extractor_class = SSD_FEATURE_EXTRACTOR_CLASS_MAP[feature_type]
return feature_extractor_class(depth_multiplier, min_depth, conv_hyperparams,
reuse_weights)
def _build_ssd_model(ssd_config, is_training):
"""Builds an SSD detection model based on the model config.
Args:
ssd_config: A ssd.proto object containing the config for the desired
SSDMetaArch.
is_training: True if this model is being built for training purposes.
Returns:
SSDMetaArch based on the config.
Raises:
ValueError: If ssd_config.type is not recognized (i.e. not registered in
model_class_map).
"""
num_classes = ssd_config.num_classes
# Feature extractor
feature_extractor = _build_ssd_feature_extractor(ssd_config.feature_extractor,
is_training)
box_coder = box_coder_builder.build(ssd_config.box_coder)
matcher = matcher_builder.build(ssd_config.matcher)
region_similarity_calculator = sim_calc.build(
ssd_config.similarity_calculator)
ssd_box_predictor = box_predictor_builder.build(hyperparams_builder.build,
ssd_config.box_predictor,
is_training, num_classes)
anchor_generator = anchor_generator_builder.build(
ssd_config.anchor_generator)
image_resizer_fn = image_resizer_builder.build(ssd_config.image_resizer)
non_max_suppression_fn, score_conversion_fn = post_processing_builder.build(
ssd_config.post_processing)
(classification_loss, localization_loss, classification_weight,
localization_weight,
hard_example_miner) = losses_builder.build(ssd_config.loss)
normalize_loss_by_num_matches = ssd_config.normalize_loss_by_num_matches
return ssd_meta_arch.SSDMetaArch(
is_training,
anchor_generator,
ssd_box_predictor,
box_coder,
feature_extractor,
matcher,
region_similarity_calculator,
image_resizer_fn,
non_max_suppression_fn,
score_conversion_fn,
classification_loss,
localization_loss,
classification_weight,
localization_weight,
normalize_loss_by_num_matches,
hard_example_miner)
def _build_faster_rcnn_feature_extractor(
feature_extractor_config, is_training, reuse_weights=None):
"""Builds a faster_rcnn_meta_arch.FasterRCNNFeatureExtractor based on config.
Args:
feature_extractor_config: A FasterRcnnFeatureExtractor proto config from
faster_rcnn.proto.
is_training: True if this feature extractor is being built for training.
reuse_weights: if the feature extractor should reuse weights.
Returns:
faster_rcnn_meta_arch.FasterRCNNFeatureExtractor based on config.
Raises:
ValueError: On invalid feature extractor type.
"""
feature_type = feature_extractor_config.type
first_stage_features_stride = (
feature_extractor_config.first_stage_features_stride)
if feature_type not in FASTER_RCNN_FEATURE_EXTRACTOR_CLASS_MAP:
raise ValueError('Unknown Faster R-CNN feature_extractor: {}'.format(
feature_type))
feature_extractor_class = FASTER_RCNN_FEATURE_EXTRACTOR_CLASS_MAP[
feature_type]
return feature_extractor_class(
is_training, first_stage_features_stride, reuse_weights)
def _build_faster_rcnn_model(frcnn_config, is_training):
"""Builds a Faster R-CNN or R-FCN detection model based on the model config.
Builds R-FCN model if the second_stage_box_predictor in the config is of type
`rfcn_box_predictor` else builds a Faster R-CNN model.
Args:
frcnn_config: A faster_rcnn.proto object containing the config for the
desired FasterRCNNMetaArch or RFCNMetaArch.
is_training: True if this model is being built for training purposes.
Returns:
FasterRCNNMetaArch based on the config.
Raises:
ValueError: If frcnn_config.type is not recognized (i.e. not registered in
model_class_map).
"""
num_classes = frcnn_config.num_classes
image_resizer_fn = image_resizer_builder.build(frcnn_config.image_resizer)
feature_extractor = _build_faster_rcnn_feature_extractor(
frcnn_config.feature_extractor, is_training)
first_stage_only = frcnn_config.first_stage_only
first_stage_anchor_generator = anchor_generator_builder.build(
frcnn_config.first_stage_anchor_generator)
first_stage_atrous_rate = frcnn_config.first_stage_atrous_rate
first_stage_box_predictor_arg_scope = hyperparams_builder.build(
frcnn_config.first_stage_box_predictor_conv_hyperparams, is_training)
first_stage_box_predictor_kernel_size = (
frcnn_config.first_stage_box_predictor_kernel_size)
first_stage_box_predictor_depth = frcnn_config.first_stage_box_predictor_depth
first_stage_minibatch_size = frcnn_config.first_stage_minibatch_size
first_stage_positive_balance_fraction = (
frcnn_config.first_stage_positive_balance_fraction)
first_stage_nms_score_threshold = frcnn_config.first_stage_nms_score_threshold
first_stage_nms_iou_threshold = frcnn_config.first_stage_nms_iou_threshold
first_stage_max_proposals = frcnn_config.first_stage_max_proposals
first_stage_loc_loss_weight = (
frcnn_config.first_stage_localization_loss_weight)
first_stage_obj_loss_weight = frcnn_config.first_stage_objectness_loss_weight
initial_crop_size = frcnn_config.initial_crop_size
maxpool_kernel_size = frcnn_config.maxpool_kernel_size
maxpool_stride = frcnn_config.maxpool_stride
second_stage_box_predictor = box_predictor_builder.build(
hyperparams_builder.build,
frcnn_config.second_stage_box_predictor,
is_training=is_training,
num_classes=num_classes)
second_stage_batch_size = frcnn_config.second_stage_batch_size
second_stage_balance_fraction = frcnn_config.second_stage_balance_fraction
(second_stage_non_max_suppression_fn, second_stage_score_conversion_fn
) = post_processing_builder.build(frcnn_config.second_stage_post_processing)
second_stage_localization_loss_weight = (
frcnn_config.second_stage_localization_loss_weight)
second_stage_classification_loss_weight = (
frcnn_config.second_stage_classification_loss_weight)
hard_example_miner = None
if frcnn_config.HasField('hard_example_miner'):
hard_example_miner = losses_builder.build_hard_example_miner(
frcnn_config.hard_example_miner,
second_stage_classification_loss_weight,
second_stage_localization_loss_weight)
common_kwargs = {
'is_training': is_training,
'num_classes': num_classes,
'image_resizer_fn': image_resizer_fn,
'feature_extractor': feature_extractor,
'first_stage_only': first_stage_only,
'first_stage_anchor_generator': first_stage_anchor_generator,
'first_stage_atrous_rate': first_stage_atrous_rate,
'first_stage_box_predictor_arg_scope':
first_stage_box_predictor_arg_scope,
'first_stage_box_predictor_kernel_size':
first_stage_box_predictor_kernel_size,
'first_stage_box_predictor_depth': first_stage_box_predictor_depth,
'first_stage_minibatch_size': first_stage_minibatch_size,
'first_stage_positive_balance_fraction':
first_stage_positive_balance_fraction,
'first_stage_nms_score_threshold': first_stage_nms_score_threshold,
'first_stage_nms_iou_threshold': first_stage_nms_iou_threshold,
'first_stage_max_proposals': first_stage_max_proposals,
'first_stage_localization_loss_weight': first_stage_loc_loss_weight,
'first_stage_objectness_loss_weight': first_stage_obj_loss_weight,
'second_stage_batch_size': second_stage_batch_size,
'second_stage_balance_fraction': second_stage_balance_fraction,
'second_stage_non_max_suppression_fn':
second_stage_non_max_suppression_fn,
'second_stage_score_conversion_fn': second_stage_score_conversion_fn,
'second_stage_localization_loss_weight':
second_stage_localization_loss_weight,
'second_stage_classification_loss_weight':
second_stage_classification_loss_weight,
'hard_example_miner': hard_example_miner}
if isinstance(second_stage_box_predictor, box_predictor.RfcnBoxPredictor):
return rfcn_meta_arch.RFCNMetaArch(
second_stage_rfcn_box_predictor=second_stage_box_predictor,
**common_kwargs)
else:
return faster_rcnn_meta_arch.FasterRCNNMetaArch(
initial_crop_size=initial_crop_size,
maxpool_kernel_size=maxpool_kernel_size,
maxpool_stride=maxpool_stride,
second_stage_mask_rcnn_box_predictor=second_stage_box_predictor,
**common_kwargs)
object_detection/builders/model_builder_test.py 0 → 100644
View file @ a4944a57
# Copyright 2017 The TensorFlow Authors. All Rights Reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
# ==============================================================================
"""Tests for object_detection.models.model_builder."""
import tensorflow as tf
from google.protobuf import text_format
from object_detection.builders import model_builder
from object_detection.meta_architectures import faster_rcnn_meta_arch
from object_detection.meta_architectures import rfcn_meta_arch
from object_detection.meta_architectures import ssd_meta_arch
from object_detection.models import faster_rcnn_inception_resnet_v2_feature_extractor as frcnn_inc_res
from object_detection.models import faster_rcnn_resnet_v1_feature_extractor as frcnn_resnet_v1
from object_detection.models.ssd_inception_v2_feature_extractor import SSDInceptionV2FeatureExtractor
from object_detection.models.ssd_mobilenet_v1_feature_extractor import SSDMobileNetV1FeatureExtractor
from object_detection.protos import model_pb2
FEATURE_EXTRACTOR_MAPS = {
'faster_rcnn_resnet50':
frcnn_resnet_v1.FasterRCNNResnet50FeatureExtractor,
'faster_rcnn_resnet101':
frcnn_resnet_v1.FasterRCNNResnet101FeatureExtractor,
'faster_rcnn_resnet152':
frcnn_resnet_v1.FasterRCNNResnet152FeatureExtractor
}
class ModelBuilderTest(tf.test.TestCase):
def create_model(self, model_config):
"""Builds a DetectionModel based on the model config.
Args:
model_config: A model.proto object containing the config for the desired
DetectionModel.
Returns:
DetectionModel based on the config.
"""
return model_builder.build(model_config, is_training=True)
def test_create_ssd_inception_v2_model_from_config(self):
model_text_proto = """
ssd {
feature_extractor {
type: 'ssd_inception_v2'
conv_hyperparams {
regularizer {
l2_regularizer {
}
}
initializer {
truncated_normal_initializer {
}
}
}
}
box_coder {
faster_rcnn_box_coder {
}
}
matcher {
argmax_matcher {
}
}
similarity_calculator {
iou_similarity {
}
}
anchor_generator {
ssd_anchor_generator {
aspect_ratios: 1.0
}
}
image_resizer {
fixed_shape_resizer {
height: 320
width: 320
}
}
box_predictor {
convolutional_box_predictor {
conv_hyperparams {
regularizer {
l2_regularizer {
}
}
initializer {
truncated_normal_initializer {
}
}
}
}
}
loss {
classification_loss {
weighted_softmax {
}
}
localization_loss {
weighted_smooth_l1 {
}
}
}
}"""
model_proto = model_pb2.DetectionModel()
text_format.Merge(model_text_proto, model_proto)
model = self.create_model(model_proto)
self.assertIsInstance(model, ssd_meta_arch.SSDMetaArch)
self.assertIsInstance(model._feature_extractor,
SSDInceptionV2FeatureExtractor)
def test_create_ssd_mobilenet_v1_model_from_config(self):
model_text_proto = """
ssd {
feature_extractor {
type: 'ssd_mobilenet_v1'
conv_hyperparams {
regularizer {
l2_regularizer {
}
}
initializer {
truncated_normal_initializer {
}
}
}
}
box_coder {
faster_rcnn_box_coder {
}
}
matcher {
argmax_matcher {
}
}
similarity_calculator {
iou_similarity {
}
}
anchor_generator {
ssd_anchor_generator {
aspect_ratios: 1.0
}
}
image_resizer {
fixed_shape_resizer {
height: 320
width: 320
}
}
box_predictor {
convolutional_box_predictor {
conv_hyperparams {
regularizer {
l2_regularizer {
}
}
initializer {
truncated_normal_initializer {
}
}
}
}
}
loss {
classification_loss {
weighted_softmax {
}
}
localization_loss {
weighted_smooth_l1 {
}
}
}
}"""
model_proto = model_pb2.DetectionModel()
text_format.Merge(model_text_proto, model_proto)
model = self.create_model(model_proto)
self.assertIsInstance(model, ssd_meta_arch.SSDMetaArch)
self.assertIsInstance(model._feature_extractor,
SSDMobileNetV1FeatureExtractor)
def test_create_faster_rcnn_resnet_v1_models_from_config(self):
model_text_proto = """
faster_rcnn {
num_classes: 3
image_resizer {
keep_aspect_ratio_resizer {
min_dimension: 600
max_dimension: 1024
}
}
feature_extractor {
type: 'faster_rcnn_resnet101'
}
first_stage_anchor_generator {
grid_anchor_generator {
scales: [0.25, 0.5, 1.0, 2.0]
aspect_ratios: [0.5, 1.0, 2.0]
height_stride: 16
width_stride: 16
}
}
first_stage_box_predictor_conv_hyperparams {
regularizer {
l2_regularizer {
}
}
initializer {
truncated_normal_initializer {
}
}
}
initial_crop_size: 14
maxpool_kernel_size: 2
maxpool_stride: 2
second_stage_box_predictor {
mask_rcnn_box_predictor {
fc_hyperparams {
op: FC
regularizer {
l2_regularizer {
}
}
initializer {
truncated_normal_initializer {
}
}
}
}
}
second_stage_post_processing {
batch_non_max_suppression {
score_threshold: 0.01
iou_threshold: 0.6
max_detections_per_class: 100
max_total_detections: 300
}
score_converter: SOFTMAX
}
}"""
model_proto = model_pb2.DetectionModel()
text_format.Merge(model_text_proto, model_proto)
for extractor_type, extractor_class in FEATURE_EXTRACTOR_MAPS.iteritems():
model_proto.faster_rcnn.feature_extractor.type = extractor_type
model = model_builder.build(model_proto, is_training=True)
self.assertIsInstance(model, faster_rcnn_meta_arch.FasterRCNNMetaArch)
self.assertIsInstance(model._feature_extractor, extractor_class)
def test_create_faster_rcnn_inception_resnet_v2_model_from_config(self):
model_text_proto = """
faster_rcnn {
num_classes: 3
image_resizer {
keep_aspect_ratio_resizer {
min_dimension: 600
max_dimension: 1024
}
}
feature_extractor {
type: 'faster_rcnn_inception_resnet_v2'
}
first_stage_anchor_generator {
grid_anchor_generator {
scales: [0.25, 0.5, 1.0, 2.0]
aspect_ratios: [0.5, 1.0, 2.0]
height_stride: 16
width_stride: 16
}
}
first_stage_box_predictor_conv_hyperparams {
regularizer {
l2_regularizer {
}
}
initializer {
truncated_normal_initializer {
}
}
}
initial_crop_size: 17
maxpool_kernel_size: 1
maxpool_stride: 1
second_stage_box_predictor {
mask_rcnn_box_predictor {
fc_hyperparams {
op: FC
regularizer {
l2_regularizer {
}
}
initializer {
truncated_normal_initializer {
}
}
}
}
}
second_stage_post_processing {
batch_non_max_suppression {
score_threshold: 0.01
iou_threshold: 0.6
max_detections_per_class: 100
max_total_detections: 300
}
score_converter: SOFTMAX
}
}"""
model_proto = model_pb2.DetectionModel()
text_format.Merge(model_text_proto, model_proto)
model = model_builder.build(model_proto, is_training=True)
self.assertIsInstance(model, faster_rcnn_meta_arch.FasterRCNNMetaArch)
self.assertIsInstance(
model._feature_extractor,
frcnn_inc_res.FasterRCNNInceptionResnetV2FeatureExtractor)
def test_create_faster_rcnn_model_from_config_with_example_miner(self):
model_text_proto = """
faster_rcnn {
num_classes: 3
feature_extractor {
type: 'faster_rcnn_inception_resnet_v2'
}
image_resizer {
keep_aspect_ratio_resizer {
min_dimension: 600
max_dimension: 1024
}
}
first_stage_anchor_generator {
grid_anchor_generator {
scales: [0.25, 0.5, 1.0, 2.0]
aspect_ratios: [0.5, 1.0, 2.0]
height_stride: 16
width_stride: 16
}
}
first_stage_box_predictor_conv_hyperparams {
regularizer {
l2_regularizer {
}
}
initializer {
truncated_normal_initializer {
}
}
}
second_stage_box_predictor {
mask_rcnn_box_predictor {
fc_hyperparams {
op: FC
regularizer {
l2_regularizer {
}
}
initializer {
truncated_normal_initializer {
}
}
}
}
}
hard_example_miner {
num_hard_examples: 10
iou_threshold: 0.99
}
}"""
model_proto = model_pb2.DetectionModel()
text_format.Merge(model_text_proto, model_proto)
model = model_builder.build(model_proto, is_training=True)
self.assertIsNotNone(model._hard_example_miner)
def test_create_rfcn_resnet_v1_model_from_config(self):
model_text_proto = """
faster_rcnn {
num_classes: 3
image_resizer {
keep_aspect_ratio_resizer {
min_dimension: 600
max_dimension: 1024
}
}
feature_extractor {
type: 'faster_rcnn_resnet101'
}
first_stage_anchor_generator {
grid_anchor_generator {
scales: [0.25, 0.5, 1.0, 2.0]
aspect_ratios: [0.5, 1.0, 2.0]
height_stride: 16
width_stride: 16
}
}
first_stage_box_predictor_conv_hyperparams {
regularizer {
l2_regularizer {
}
}
initializer {
truncated_normal_initializer {
}
}
}
initial_crop_size: 14
maxpool_kernel_size: 2
maxpool_stride: 2
second_stage_box_predictor {
rfcn_box_predictor {
conv_hyperparams {
op: CONV
regularizer {
l2_regularizer {
}
}
initializer {
truncated_normal_initializer {
}
}
}
}
}
second_stage_post_processing {
batch_non_max_suppression {
score_threshold: 0.01
iou_threshold: 0.6
max_detections_per_class: 100
max_total_detections: 300
}
score_converter: SOFTMAX
}
}"""
model_proto = model_pb2.DetectionModel()
text_format.Merge(model_text_proto, model_proto)
for extractor_type, extractor_class in FEATURE_EXTRACTOR_MAPS.iteritems():
model_proto.faster_rcnn.feature_extractor.type = extractor_type
model = model_builder.build(model_proto, is_training=True)
self.assertIsInstance(model, rfcn_meta_arch.RFCNMetaArch)
self.assertIsInstance(model._feature_extractor, extractor_class)
if __name__ == '__main__':
tf.test.main()
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