def test_merge_dict(self):
target_dict = {'a': 1, 'b': 2}
source_dict = {'c': 4}
data_utils.merge_dict(source_dict, target_dict)
self.assertDictEqual(target_dict, {'a': 1, 'b': 2, 'c': 4})
target_dict = {'a': 1, 'b': 2}
source_dict = {'b': 3, 'c': 4}
with self.assertRaisesRegexp(ValueError, 'Key conflict: `b`.'):
data_utils.merge_dict(source_dict, target_dict)
def create_inputs():
"""Creates pipeline and model inputs."""
inputs = pipeline_utils.read_batch_from_dataset_tables(
FLAGS.input_table,
batch_sizes=[int(x) for x in FLAGS.batch_size],
num_instances_per_record=2,
shuffle=True,
num_epochs=None,
keypoint_names_3d=configs['keypoint_profile_3d'].
keypoint_names,
keypoint_names_2d=configs['keypoint_profile_2d'].
keypoint_names,
min_keypoint_score_2d=FLAGS.min_input_keypoint_score_2d,
shuffle_buffer_size=FLAGS.input_shuffle_buffer_size,
common_module=common_module,
dataset_class=input_dataset_class,
input_example_parser_creator=input_example_parser_creator)
(inputs[common_module.KEY_KEYPOINTS_3D],
keypoint_preprocessor_side_outputs_3d
) = keypoint_preprocessor_3d(
inputs[common_module.KEY_KEYPOINTS_3D],
keypoint_profile_3d=configs['keypoint_profile_3d'],
normalize_keypoints_3d=True)
inputs.update(keypoint_preprocessor_side_outputs_3d)
inputs['model_inputs'], side_inputs = configs[
'create_model_input_fn'](
inputs[common_module.KEY_KEYPOINTS_2D],
inputs[common_module.KEY_KEYPOINT_MASKS_2D],
inputs[common_module.KEY_PREPROCESSED_KEYPOINTS_3D],
model_input_keypoint_type=FLAGS.
model_input_keypoint_type,
normalize_keypoints_2d=True,
keypoint_profile_2d=configs['keypoint_profile_2d'],
keypoint_profile_3d=configs['keypoint_profile_3d'],
azimuth_range=configs[
'random_projection_azimuth_range'],
elevation_range=configs[
'random_projection_elevation_range'],
roll_range=configs['random_projection_roll_range'],
normalized_camera_depth_range=(
configs['random_projection_camera_depth_range']))
data_utils.merge_dict(side_inputs, inputs)
return inputs