openvino/model-optimizer/extensions/analysis/tf_yolo.py

121 lines
5.1 KiB
Python

"""
Copyright (C) 2018-2020 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.
"""
from mo.graph.graph import Graph
from mo.middle.pattern_match import apply_pattern
from mo.utils.model_analysis import AnalyzeAction, graph_contains_scope
YOLO_PATTERN = {
'nodes': [
('pad', dict(op='Pad')),
('conv', dict(op='Conv2D')),
('sub', dict(op='Sub')),
('div', dict(op='Div')),
('mul', dict(op='Mul')),
('bias_add', dict(op='Add')),
('mul_2', dict(op='Mul')),
('max', dict(op='Maximum')),
],
'edges': [
('pad', 'conv', {'out': 0}),
('conv', 'sub', {'out': 0}),
('sub', 'div', {'out': 0}),
('div', 'mul', {'out': 0}),
('mul', 'bias_add', {'out': 0}),
('bias_add', 'mul_2', {'out': 0}),
('bias_add', 'max', {'out': 0}),
('mul_2', 'max', {'out': 0}),
]
}
def pattern_instance_counter(graph: Graph, match: dict):
pattern_instance_counter.counter += 1
pattern_instance_counter.counter = 0
YOLO_CONFIGS = {'YOLOV2Full': ['extensions/front/tf/yolo_v2.json', 'extensions/front/tf/yolo_v2_voc.json'],
'YOLOV3Full': ['extensions/front/tf/yolo_v3.json', 'extensions/front/tf/yolo_v3_voc.json'],
'YOLOV2Tiny': ['extensions/front/tf/yolo_v2_tiny.json', 'extensions/front/tf/yolo_v2_tiny_voc.json'],
'YOLOV3Tiny': ['extensions/front/tf/yolo_v3_tiny.json', 'extensions/front/tf/yolo_v3_tiny_voc.json'],
}
def get_YOLO_params_by_flavor(flavor: str):
result = dict()
result['flavor'] = flavor
result['mandatory_parameters'] = {'tensorflow_use_custom_operations_config': YOLO_CONFIGS[flavor]}
return result
class TensorFlowYOLOV1V2Analysis(AnalyzeAction):
"""
The analyser checks if the provided model is TensorFlow YOLO models from https://github.com/thtrieu/darkflow .
"""
graph_condition = [lambda graph: graph.graph['fw'] == 'tf']
def analyze(self, graph: Graph):
pattern_instance_counter.counter = 0
apply_pattern(graph, **YOLO_PATTERN, action=pattern_instance_counter)
flavor = None
if pattern_instance_counter.counter > 0:
if pattern_instance_counter.counter == 22:
flavor = 'YOLOV2Full'
elif pattern_instance_counter.counter == 8:
flavor = 'YOLOV2Tiny'
if flavor is not None:
message = "Your model looks like YOLOv1 or YOLOv2 Model.\n" \
"To generate the IR, provide TensorFlow YOLOv1 or YOLOv2 Model to the Model Optimizer with the following parameters:\n" \
"\t--input_model <path_to_model>/<model_name>.pb\n" \
"\t--batch 1\n" \
"\t--tensorflow_use_custom_operations_config <OPENVINO_INSTALL_DIR>/deployment_tools/model_optimizer/extensions/front/tf/<yolo_config>.json\n" \
"All detailed information about conversion of this model can be found at\n" \
"https://docs.openvinotoolkit.org/latest/_docs_MO_DG_prepare_model_convert_model_tf_specific_Convert_YOLO_From_Tensorflow.html"
return {'model_type': {'YOLO': get_YOLO_params_by_flavor(flavor)}}, message
else:
return None, None
class TensorFlowYOLOV3Analysis(AnalyzeAction):
"""
The analyser checks if the provided model is TensorFlow YOLO models from
https://github.com/mystic123/tensorflow-yolo-v3.
"""
graph_condition = [lambda graph: graph.graph['fw'] == 'tf']
def analyze(self, graph: Graph):
flavor = None
if graph_contains_scope(graph, 'detector/yolo-v3') and graph_contains_scope(graph, 'detector/darknet-53'):
flavor = 'YOLOV3Full'
elif graph_contains_scope(graph, 'detector/yolo-v3-tiny'):
flavor = 'YOLOV3Tiny'
if flavor is not None:
message = "Your model looks like YOLOv3 Model.\n" \
"To generate the IR, provide TensorFlow YOLOv3 Model to the Model Optimizer with the following parameters:\n" \
"\t--input_model <path_to_model>/yolo_v3.pb\n" \
"\t--batch 1\n" \
"\t--tensorflow_use_custom_operations_config <OPENVINO_INSTALL_DIR>/deployment_tools/model_optimizer/extensions/front/tf/yolo_v3.json\n" \
"Detailed information about conversion of this model can be found at\n" \
"https://docs.openvinotoolkit.org/latest/_docs_MO_DG_prepare_model_convert_model_tf_specific_Convert_YOLO_From_Tensorflow.html"
return {'model_type': {'YOLO': get_YOLO_params_by_flavor(flavor)}}, message
else:
return None, None