Intel® Distribution of OpenVINO™ Toolkit
Community support and discussions about the Intel® Distribution of OpenVINO™ toolkit, OpenCV, and all things computer vision-related on Intel® platforms.
6023 Discussions

Illegal Instruction when using Model Optimizer (Robot Devkit installation)



I am currently working on installing the Intel robot devkit on an UP squared board with a fresh install of Ubuntu 18.04 LTS. (I got an Intel RealSense D435i as well, but not connected yet.)

All I have done so far is installing git and protobuf and then downloading and installing the sdk as described here:

At one point in the installation routine (as part of the openvino installation), the tensorflow modeloptimizer throws an 'Illegal Instruction' which I can not solve.


Thanks for your support!






Model Optimizer arguments:
Common parameters:
    - Path to the Input Model:     /home/ifu/Downloads/models/mask_rcnn_inception_v2_coco_2018_01_28/frozen_inference_graph.pb
    - Path for generated IR:     /opt/openvino_toolkit/models/segmentation/output/FP32
    - IR output name:     frozen_inference_graph
    - Log level:     ERROR
    - 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:     Not specified, inherited from the model
    - 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:     True
TensorFlow specific parameters:
    - Input model in text protobuf format:     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:     /home/ifu/Downloads/models/mask_rcnn_inception_v2_coco_2018_01_28/pipeline.config
    - Operations to offload:     None
    - Patterns to offload:     None
    - Use the config file:     /opt/intel/openvino/deployment_tools/model_optimizer/extensions/front/tf/mask_rcnn_support.json
Model Optimizer version:     2019.1.1-83-g28dfbfd
/home/ifu/robot_devkit/packages/perception/deps/33-openvino.deps: line 14: 30707 Invalid instruction   sudo python3 /opt/intel/openvino/deployment_tools/model_optimizer/ --input_model frozen_inference_graph.pb --tensorflow_use_custom_operations_config /opt/intel/openvino/deployment_tools/model_optimizer/extensions/front/tf/mask_rcnn_support.json --tensorflow_object_detection_api_pipeline_config pipeline.config --reverse_input_channels --output_dir /opt/openvino_toolkit/models/segmentation/output/FP32






Labels (1)
0 Kudos
6 Replies
Community Manager

Hi Ifumho,


Thanks for reaching out.

We are currently investigating this issue and will get back to you. Which OpenVINO version you are using? Can you please share your frozen graph for us to reproduce the workaround?





Hi Aznie,

thanks for your initial message!

My OpenVINO version is 2019.1.144.
The frozen graph should be attached (I hope, this is, what you need).
Just tell me, if you need more information.

And for reproduction:
My system is the UP Squared Board (Intel Celeron N3350, 4 GB DDR4). I used a fresh install of Ubuntu 18.04.5 Desktop (Kernel 5.4.0-53-generic, but others tested as well) and only executed the following commands afterwards:

sudo apt-get install git libprotobuf-dev protobuf-compiler
git clone
cd robot_devkit


Thanks for your help!

Community Manager

 Hi Mario,

Can I know how you installing the OpenVINO toolkit? Did you refer to the official OpenVINO toolkit installation documentation? If not, please follow the installation step from the following link.

Also, make sure your system meets the OpenVINO installation requirement. The OpenVINO 2019 version was only supported on Ubuntu 16.04 LTS.


Therefore, it is recommended to use the latest version of the OpenVINO toolkit which is the OpenVINO 2021.1. The latest version is more stable and upgraded.





Hi Aznie,

I did NOT install OpenVINO manually. As described above, I have just used the installation routine from the Intel Robot Devkit, which automatically installs OpenVINO. And since the Github page ( states, that only Ubuntu 18.04 is supported, I used this OS.

Therefore, I did not choose to install OpenVINO manually but was "forced" by the Intel Robot DevKit installer to just use this specific version. When manually switching to a newer version, I doubt, that the Robot DevKit will work with that. If there is a chance to use the Robot DevKit with OpenVINO 2021, I would be happy if you can provide me with information on how to do that.

Best regards,


Community Manager

Hi Mario,


The Robot Devkit is an open-source project and not officially validated by OpenVINO Toolkit. Also, the Robot Devkit project was designed quite a while ago and no new version is released, so we should deal with what we have now. It is recommended for you to try installing the latest version of OpenVINO Toolkit but we cannot guarantee the functionality within the project.




Community Manager

Hi Mario,

This thread will no longer be monitored since we have provided a solution. If you need any additional information from Intel, please submit a new question.