diff options
Diffstat (limited to 'model-optimizer/extensions/front/tf/Preprocessor.py')
| -rw-r--r-- | model-optimizer/extensions/front/tf/Preprocessor.py | 122 |
1 files changed, 0 insertions, 122 deletions
diff --git a/model-optimizer/extensions/front/tf/Preprocessor.py b/model-optimizer/extensions/front/tf/Preprocessor.py deleted file mode 100644 index 207f28851..000000000 --- a/model-optimizer/extensions/front/tf/Preprocessor.py +++ /dev/null @@ -1,122 +0,0 @@ -""" - Copyright (c) 2018 Intel Corporation - - 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. -""" - -import logging as log -import networkx as nx - -from extensions.front.sub import Sub -from extensions.front.tf.Pack import Pack -from mo.front.subgraph_matcher import SubgraphMatch -from mo.front.tf.replacement import FrontReplacementFromConfigFileSubGraph -from mo.graph.graph import create_edge, Node -from mo.utils.error import Error - - -class PreprocessorReplacement(FrontReplacementFromConfigFileSubGraph): - """ - The class replaces the "Preprocessor" block resizing input image and applying mean/scale values. Only nodes related - to applying mean/scaling values are kept. - """ - replacement_id = 'PreprocessorReplacement' - - def run_before(self): - return [Pack, Sub] - - def nodes_to_remove(self, graph: nx.MultiDiGraph, match: SubgraphMatch): - new_nodes_to_remove = match.matched_nodes_names() - # do not remove nodes that perform input image scaling and mean value subtraction - for node_to_keep in ('Preprocessor/sub', 'Preprocessor/sub/y', 'Preprocessor/mul', 'Preprocessor/mul/x'): - if node_to_keep in new_nodes_to_remove: - new_nodes_to_remove.remove(node_to_keep) - return new_nodes_to_remove - - def generate_sub_graph(self, graph: nx.MultiDiGraph, match: SubgraphMatch): - print('WARNING: the "{}" is a legacy replacer that will be removed in the future release. Please, consider ' - 'using replacers defined in the "extensions/front/tf/ObjectDetectionAPI.py"'.format(self.replacement_id)) - log.debug('PreprocessorReplacement: matched_nodes = {}'.format(match.matched_nodes_names())) - - sub_node = match.output_node(0)[0] - if not sub_node.has('op') or sub_node.op != 'Sub': - raise Error('The output op of the Preprocessor sub-graph is not of type "Sub". Looks like the topology is ' - 'not created with TensorFlow Object Detection API.') - - mul_node = None - if sub_node.in_node(0).has('op') and sub_node.in_node(0).op == 'Mul': - log.info('There is image scaling node in the Preprocessor block.') - mul_node = sub_node.in_node(0) - - config_attrs = match.custom_replacement_desc.custom_attributes - preprocessed_image_height_width = self.get_preprocessed_image_size_from_model(graph) - if preprocessed_image_height_width is None: - if 'preprocessed_image_width' not in config_attrs or 'preprocessed_image_height' not in config_attrs: - raise Error('Failed to determine the pre-processed image size from the original TensorFlow graph. ' - 'Please, specify "preprocessed_image_width" and "preprocessed_image_height" in the ' - 'topology replacement configuration file in the "custom_attributes" section of the ' - '"PreprocessorReplacement" replacer. This value is defined in the configuration file ' - 'samples/configs/*.config of the model in the Object Detection model zoo as ' - '"min_dimension".') - else: - graph.graph['preprocessed_image_width'] = config_attrs['preprocessed_image_width'] - graph.graph['preprocessed_image_height'] = config_attrs['preprocessed_image_height'] - else: - graph.graph['preprocessed_image_height'] = preprocessed_image_height_width[0] - graph.graph['preprocessed_image_width'] = preprocessed_image_height_width[1] - - initial_input_node_name = 'image_tensor' - if initial_input_node_name not in graph.nodes(): - raise Error('Input node "{}" of the graph is not found. Do not run the Model Optimizer with ' - '"--input" command line parameter.'.format(initial_input_node_name)) - placeholder_node = Node(graph, initial_input_node_name) - - if placeholder_node.shape[0] != 1 and placeholder_node.shape[0] != -1: - raise Error('The faster R-CNN model support batch size 1 only.') - placeholder_node.shape[0] = 1 # batch size 1 is supported only - placeholder_node.shape[1] = graph.graph['preprocessed_image_height'] - placeholder_node.shape[2] = graph.graph['preprocessed_image_width'] - - to_float_node = placeholder_node.out_node(0) - if not to_float_node.has('op') or to_float_node.op != 'Cast': - raise Error('The output of the "{}" is not Cast operation. Cannot apply replacer.'.format( - initial_input_node_name)) - - # connect to_float_node directly with node performing scale on mean value subtraction - if mul_node is None: - create_edge(to_float_node, sub_node, 0, 0) - else: - create_edge(to_float_node, mul_node, 0, 1) - - print('The Preprocessor block has been removed. Only nodes performing mean value subtraction and scaling (if' - ' applicable) are kept.') - return {} - - @staticmethod - def get_preprocessed_image_size_from_model(graph: nx.MultiDiGraph): - """ - The function looks for nodes in the Preprocessor block with specific names for resized image shape. If one of - the nodes exist return the desired size. If nodes do not exist then return None. - :param graph: graph to operate on. - :return: the tuple with height and width of the preprocessed image. - """ - preprocess_resize_to_range_size_node_name = 'Preprocessor/map/while/ResizeToRange/Const' - preprocess_resize_bilinear_node_name = 'Preprocessor/map/while/ResizeImage/ResizeBilinear' - result = None - if preprocess_resize_to_range_size_node_name in graph.nodes(): - preprocess_size_node = Node(graph, preprocess_resize_to_range_size_node_name) - result = (int(preprocess_size_node.value.item()), int(preprocess_size_node.value.item())) - elif preprocess_resize_bilinear_node_name in graph.nodes(): - preprocess_size_node = Node(graph, preprocess_resize_bilinear_node_name) - result = (int(preprocess_size_node.in_node(1).value[0]), int(preprocess_size_node.in_node(1).value[1])) - return result |