Python google.protobuf.text_format.Merge() Examples

def testBuildMomentumOptimizer(self):
    optimizer_text_proto = """
      momentum_optimizer: {
        learning_rate: {
          constant_learning_rate {
            learning_rate: 0.001
          }
        }
        momentum_optimizer_value: 0.99
      }
      use_moving_average: false
    """
    global_summaries = set([])
    optimizer_proto = optimizer_pb2.Optimizer()
    text_format.Merge(optimizer_text_proto, optimizer_proto)
    optimizer = optimizer_builder.build(optimizer_proto, global_summaries)
    self.assertTrue(isinstance(optimizer, tf.train.MomentumOptimizer)) 

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 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_build_non_max_suppressor_with_correct_parameters(self):
    post_processing_text_proto = """
      batch_non_max_suppression {
        score_threshold: 0.7
        iou_threshold: 0.6
        max_detections_per_class: 100
        max_total_detections: 300
      }
    """
    post_processing_config = post_processing_pb2.PostProcessing()
    text_format.Merge(post_processing_text_proto, post_processing_config)
    non_max_suppressor, _ = post_processing_builder.build(
        post_processing_config)
    self.assertEqual(non_max_suppressor.keywords['max_size_per_class'], 100)
    self.assertEqual(non_max_suppressor.keywords['max_total_size'], 300)
    self.assertAlmostEqual(non_max_suppressor.keywords['score_thresh'], 0.7)
    self.assertAlmostEqual(non_max_suppressor.keywords['iou_thresh'], 0.6) 

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_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) 

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_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 testBuildManualStepLearningRate(self):
    learning_rate_text_proto = """
      manual_step_learning_rate {
        schedule {
          step: 0
          learning_rate: 0.006
        }
        schedule {
          step: 90000
          learning_rate: 0.00006
        }
      }
    """
    global_summaries = set([])
    learning_rate_proto = optimizer_pb2.LearningRate()
    text_format.Merge(learning_rate_text_proto, learning_rate_proto)
    learning_rate = optimizer_builder._create_learning_rate(
        learning_rate_proto, global_summaries)
    self.assertTrue(isinstance(learning_rate, tf.Tensor)) 

def testBuildAdamOptimizer(self):
    optimizer_text_proto = """
      adam_optimizer: {
        learning_rate: {
          constant_learning_rate {
            learning_rate: 0.002
          }
        }
      }
      use_moving_average: false
    """
    global_summaries = set([])
    optimizer_proto = optimizer_pb2.Optimizer()
    text_format.Merge(optimizer_text_proto, optimizer_proto)
    optimizer = optimizer_builder.build(optimizer_proto, global_summaries)
    self.assertTrue(isinstance(optimizer, tf.train.AdamOptimizer)) 

def get_configs_from_pipeline_file():
  """Reads evaluation configuration from a pipeline_pb2.TrainEvalPipelineConfig.

  Reads evaluation config from file specified by pipeline_config_path flag.

  Returns:
    model_config: a model_pb2.DetectionModel
    eval_config: a eval_pb2.EvalConfig
    input_config: a input_reader_pb2.InputReader
  """
  pipeline_config = pipeline_pb2.TrainEvalPipelineConfig()
  with tf.gfile.GFile(FLAGS.pipeline_config_path, 'r') as f:
    text_format.Merge(f.read(), pipeline_config)

  model_config = pipeline_config.model
  if FLAGS.eval_training_data:
    eval_config = pipeline_config.train_config
  else:
    eval_config = pipeline_config.eval_config
  input_config = pipeline_config.eval_input_reader

  return model_config, eval_config, input_config 

def testBuildMovingAverageOptimizer(self):
    optimizer_text_proto = """
      adam_optimizer: {
        learning_rate: {
          constant_learning_rate {
            learning_rate: 0.002
          }
        }
      }
      use_moving_average: True
    """
    global_summaries = set([])
    optimizer_proto = optimizer_pb2.Optimizer()
    text_format.Merge(optimizer_text_proto, optimizer_proto)
    optimizer = optimizer_builder.build(optimizer_proto, global_summaries)
    self.assertTrue(
        isinstance(optimizer, tf.contrib.opt.MovingAverageOptimizer)) 

def testBuildMovingAverageOptimizerWithNonDefaultDecay(self):
    optimizer_text_proto = """
      adam_optimizer: {
        learning_rate: {
          constant_learning_rate {
            learning_rate: 0.002
          }
        }
      }
      use_moving_average: True
      moving_average_decay: 0.2
    """
    global_summaries = set([])
    optimizer_proto = optimizer_pb2.Optimizer()
    text_format.Merge(optimizer_text_proto, optimizer_proto)
    optimizer = optimizer_builder.build(optimizer_proto, global_summaries)
    self.assertTrue(
        isinstance(optimizer, tf.contrib.opt.MovingAverageOptimizer))
    # TODO: Find a way to not depend on the private members.
    self.assertAlmostEqual(optimizer._ema._decay, 0.2) 

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 load_labelmap(path):
  """Loads label map proto.

  Args:
    path: path to StringIntLabelMap proto text file.
  Returns:
    a StringIntLabelMapProto
  """
  with tf.gfile.GFile(path, 'r') as fid:
    label_map_string = fid.read()
    label_map = string_int_label_map_pb2.StringIntLabelMap()
    try:
      text_format.Merge(label_map_string, label_map)
    except text_format.ParseError:
      label_map.ParseFromString(label_map_string)
  _validate_label_map(label_map)
  return label_map 

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_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_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_build_normalize_image(self):
    preprocessor_text_proto = """
    normalize_image {
      original_minval: 0.0
      original_maxval: 255.0
      target_minval: -1.0
      target_maxval: 1.0
    }
    """
    preprocessor_proto = preprocessor_pb2.PreprocessingStep()
    text_format.Merge(preprocessor_text_proto, preprocessor_proto)
    function, args = preprocessor_builder.build(preprocessor_proto)
    self.assertEqual(function, preprocessor.normalize_image)
    self.assertEqual(args, {
        'original_minval': 0.0,
        'original_maxval': 255.0,
        'target_minval': -1.0,
        'target_maxval': 1.0,
    }) 

def test_build_random_crop_to_aspect_ratio(self):
    preprocessor_text_proto = """
    random_crop_to_aspect_ratio {
      aspect_ratio: 0.85
      overlap_thresh: 0.35
    }
    """
    preprocessor_proto = preprocessor_pb2.PreprocessingStep()
    text_format.Merge(preprocessor_text_proto, preprocessor_proto)
    function, args = preprocessor_builder.build(preprocessor_proto)
    self.assertEqual(function, preprocessor.random_crop_to_aspect_ratio)
    self.assert_dictionary_close(args, {'aspect_ratio': 0.85,
                                        'overlap_thresh': 0.35}) 

def test_build_resize_image(self):
    preprocessor_text_proto = """
    resize_image {
      new_height: 75
      new_width: 100
      method: BICUBIC
    }
    """
    preprocessor_proto = preprocessor_pb2.PreprocessingStep()
    text_format.Merge(preprocessor_text_proto, preprocessor_proto)
    function, args = preprocessor_builder.build(preprocessor_proto)
    self.assertEqual(function, preprocessor.resize_image)
    self.assertEqual(args, {'new_height': 75,
                            'new_width': 100,
                            'method': tf.image.ResizeMethod.BICUBIC}) 

def test_build_random_black_patches(self):
    preprocessor_text_proto = """
    random_black_patches {
      max_black_patches: 20
      probability: 0.95
      size_to_image_ratio: 0.12
    }
    """
    preprocessor_proto = preprocessor_pb2.PreprocessingStep()
    text_format.Merge(preprocessor_text_proto, preprocessor_proto)
    function, args = preprocessor_builder.build(preprocessor_proto)
    self.assertEqual(function, preprocessor.random_black_patches)
    self.assert_dictionary_close(args, {'max_black_patches': 20,
                                        'probability': 0.95,
                                        'size_to_image_ratio': 0.12}) 

def test_build_random_crop_pad_image(self):
    preprocessor_text_proto = """
    random_crop_pad_image {
      min_object_covered: 0.75
      min_aspect_ratio: 0.75
      max_aspect_ratio: 1.5
      min_area: 0.25
      max_area: 0.875
      overlap_thresh: 0.5
      random_coef: 0.125
    }
    """
    preprocessor_proto = preprocessor_pb2.PreprocessingStep()
    text_format.Merge(preprocessor_text_proto, preprocessor_proto)
    function, args = preprocessor_builder.build(preprocessor_proto)
    self.assertEqual(function, preprocessor.random_crop_pad_image)
    self.assertEqual(args, {
        'min_object_covered': 0.75,
        'aspect_ratio_range': (0.75, 1.5),
        'area_range': (0.25, 0.875),
        'overlap_thresh': 0.5,
        'random_coef': 0.125,
        'min_padded_size_ratio': None,
        'max_padded_size_ratio': None,
        'pad_color': None,
    }) 

def test_build_ssd_random_crop(self):
    preprocessor_text_proto = """
    ssd_random_crop {
      operations {
        min_object_covered: 0.0
        min_aspect_ratio: 0.875
        max_aspect_ratio: 1.125
        min_area: 0.5
        max_area: 1.0
        overlap_thresh: 0.0
        random_coef: 0.375
      }
      operations {
        min_object_covered: 0.25
        min_aspect_ratio: 0.75
        max_aspect_ratio: 1.5
        min_area: 0.5
        max_area: 1.0
        overlap_thresh: 0.25
        random_coef: 0.375
      }
    }
    """
    preprocessor_proto = preprocessor_pb2.PreprocessingStep()
    text_format.Merge(preprocessor_text_proto, preprocessor_proto)
    function, args = preprocessor_builder.build(preprocessor_proto)
    self.assertEqual(function, preprocessor.ssd_random_crop)
    self.assertEqual(args, {'min_object_covered': [0.0, 0.25],
                            'aspect_ratio_range': [(0.875, 1.125), (0.75, 1.5)],
                            'area_range': [(0.5, 1.0), (0.5, 1.0)],
                            'overlap_thresh': [0.0, 0.25],
                            'random_coef': [0.375, 0.375]}) 

def test_build_random_jitter_boxes(self):
    preprocessor_text_proto = """
    random_jitter_boxes {
      ratio: 0.1
    }
    """
    preprocessor_proto = preprocessor_pb2.PreprocessingStep()
    text_format.Merge(preprocessor_text_proto, preprocessor_proto)
    function, args = preprocessor_builder.build(preprocessor_proto)
    self.assertEqual(function, preprocessor.random_jitter_boxes)
    self.assert_dictionary_close(args, {'ratio': 0.1}) 

def test_build_random_resize_method(self):
    preprocessor_text_proto = """
    random_resize_method {
      target_height: 75
      target_width: 100
    }
    """
    preprocessor_proto = preprocessor_pb2.PreprocessingStep()
    text_format.Merge(preprocessor_text_proto, preprocessor_proto)
    function, args = preprocessor_builder.build(preprocessor_proto)
    self.assertEqual(function, preprocessor.random_resize_method)
    self.assert_dictionary_close(args, {'target_size': [75, 100]}) 

def test_build_ssd_random_crop_empty_operations(self):
    preprocessor_text_proto = """
    ssd_random_crop {
    }
    """
    preprocessor_proto = preprocessor_pb2.PreprocessingStep()
    text_format.Merge(preprocessor_text_proto, preprocessor_proto)
    function, args = preprocessor_builder.build(preprocessor_proto)
    self.assertEqual(function, preprocessor.ssd_random_crop)
    self.assertEqual(args, {}) 

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)