OpenVino Toolkit Model Optimizer Error with Strided Slice Layer

[Affolter__Martin]

Hello, I get this internal error when running the model optimizer on a Tensorflow model. The whole output is attached as a file, I can also provide the model if needed. I have seen in the forum that the optimizer has problems with strided slice layers, but have not found a solution. How can I get successfully create an IR of the model? Error output:

[ ERROR ]  -------------------------------------------------
[ ERROR ]  ----------------- INTERNAL ERROR ----------------
[ ERROR ]  Unexpected exception happened.
[ ERROR ]  Please contact Model Optimizer developers and forward the following information:
[ ERROR ]  0
[ ERROR ]  Traceback (most recent call last):
  File "/opt/intel/computer_vision_sdk_2018.5.445/deployment_tools/model_optimizer/mo/main.py", line 325, in main
    return driver(argv)
  File "/opt/intel/computer_vision_sdk_2018.5.445/deployment_tools/model_optimizer/mo/main.py", line 267, in driver
    mean_scale_values=mean_scale)
  File "/opt/intel/computer_vision_sdk_2018.5.445/deployment_tools/model_optimizer/mo/pipeline/tf.py", line 297, in tf2nx
    replacer.find_and_replace_pattern(graph)
  File "/opt/intel/computer_vision_sdk_2018.5.445/deployment_tools/model_optimizer/mo/utils/replacement_pattern.py", line 28, in find_and_replace_pattern
    apply_pattern(graph, **self.pattern(), action=self.replace_pattern)  # pylint: disable=no-member
  File "/opt/intel/computer_vision_sdk_2018.5.445/deployment_tools/model_optimizer/mo/middle/pattern_match.py", line 95, in apply_pattern
    action(graph, match)
  File "/opt/intel/computer_vision_sdk_2018.5.445/deployment_tools/model_optimizer/extensions/middle/TensorIteratorMerge.py", line 103, in replace_pattern
    cond_data = match['condition'].out_node(0)
  File "/opt/intel/computer_vision_sdk_2018.5.445/deployment_tools/model_optimizer/mo/graph/graph.py", line 332, in out_node
    return self.out_nodes(control_flow=control_flow)[key]
KeyError: 0

[ ERROR ]  ---------------- END OF BUG REPORT --------------
[ ERROR ]  -------------------------------------------------

Model Optimizer arguments:
Common parameters:
	- Path to the Input Model: 	/media/amitrevda/storage/openvino_toolkit/input_models/juno_body_appearance/mars-small128.pb
	- Path for generated IR: 	/media/amitrevda/storage/openvino_toolkit/output_models/
	- IR output name: 	mars-small128
	- Log level: 	DEBUG
	- Batch: 	Not specified, inherited from the model
	- Input layers: 	Not specified, inherited from the model
	- Output layers: 	Not specified, inherited from the model
	- Input shapes: 	[1,128,64,3]
	- Mean values: 	Not specified
	- Scale values: 	Not specified
	- Scale factor: 	Not specified
	- Precision of IR: 	FP32
	- Enable fusing: 	True
	- Enable grouped convolutions fusing: 	True
	- Move mean values to preprocess section: 	False
	- Reverse input channels: 	False
TensorFlow specific parameters:
	- Input model in text protobuf format: 	False
	- Offload unsupported operations: 	False
	- Path to model dump for TensorBoard: 	None
	- List of shared libraries with TensorFlow custom layers implementation: 	None
	- Update the configuration file with input/output node names: 	None
	- Use configuration file used to generate the model with Object Detection API: 	None
	- Operations to offload: 	None
	- Patterns to offload: 	None
	- Use the config file: 	None
Model Optimizer version: 	1.5.12.49d067a0

[Severine_H_Intel]

Dear Martin, 

which topology are you using? Can you send your model? (use the PM function). I will investigate this issue and escalate to the development team if necessary.

Best, 

Severine

[Affolter__Martin]

Dear Severine,

Thanks for your answer. The model we've tried to convert can be found here: https://github.com/nwojke/deep_sort
(look for "Then, download pre-generated detections and the CNN checkpoint file from here.")
It is a Wide Residual Network, the CNN architecture is described in the following paper: https://arxiv.org/abs/1703.07402

Best,
Martin

PS: This problem has become irrelevant for us. I am providing this answer in case you are interested.

[Singh__Aditya]

Habert, Severine (Intel) wrote:

Dear Martin, 

which topology are you using? Can you send your model? (use the PM function). I will investigate this issue and escalate to the development team if necessary.

Best, 

Severine

Any updates on the issue? 

[倪__嘉旻]

Dear Martin, 

l have the same problem as yours.

So have you solved your problem?

Best regards,

Kathryn

[倪__嘉旻]

Habert, Severine (Intel) wrote:

Dear Martin, 

which topology are you using? Can you send your model? (use the PM function). I will investigate this issue and escalate to the development team if necessary.

Best, 

Severine

Dear Severine，

l meet with the same problem.

can you help me to solve it?

Thanks a lot!!

Best regards,

Kathryn