[Bug]RuntimeError: Device with "CPU" name is not recognised in InferenceEngine

[MiraGaikwad]

System information (version)

OpenVINO=I am able to make inference using Raspberry Pi 3 + Neural Compute Stick2 but unable to make inference using Raspberry Pi3 + CPU

OpenVINO = 2021.4.2 Operating System / Platform => Raspeberry Pi 3 , version 10 (buster) Problem classification = Inference Python Version = 3.7

Detailed description

I am able to make inference using Raspberry Pi 3 + Neural Compute Stick2 but unable to make inference using Raspberry Pi3 + CPU

Device with "CPU" name is not recognised in Inference Engine

Steps to reproduce

Running model with command python3 hello_classification.py -m IR_model/IR_model.xml -i <image_name.tif> -d CPU

Issue submission checklist

• [ ] I report the issue, it's not a question
• [ ] I checked the problem with documentation, FAQ, open issues, Stack Overflow, etc and have not found solution
• [ ] There is reproducer code and related data files: images, videos, models, etc.

[brmarkus]

Please have a look at the ARM plugin, e.g. here: https://github.com/openvinotoolkit/openvino_contrib/blob/master/modules/arm_plugin/README.md

[MiraGaikwad]

@brmarkus - Thanks for the response. I checked the link and looks like those plugins are for Raspberry Pi4 but I am using Raspberry pi 3.

Do you know if plugins are available for Raspberry Pi3 as well ?

Cheers Mira

[brmarkus]

Sorry, cannot answer, I haven't used the ARM plugins on my own, yet. You might need to seek for help under "https://github.com/openvinotoolkit/openvino_contrib".

[Iffa-Intel]

I had tried to link that arm CPU plugin from openvino_contrib with OpenVINO build from the source but no luck so far.

command: cmake -DInferenceEngineDeveloperPackage_DIR=~/Downloads/openvino/build -DCMAKE_BUILD_TYPE=Release .. && make

Only MYRIAD is listed.

We'll get back to you with possible workaround soon. Thank you for your patience.

[MiraGaikwad]

Thanks al ot @Iffa-Meah ..

@brmarkus , @Iffa-Meah -

Meanwhile, I have brought Raspberry Pi4 Model B 8 GB RAM. OpenVINO = 2021.4.2 Operating System / Platform => Raspeberry Pi 4 64 bit, version 10 (buster) Problem classification = Inference Python Version = 3.7

I am still getting same error.. Do I need to install the Plugins separately and if yes could you please let me know the process to download it ?

Also, could you Please let me know if I need to upgrade the OS as I am currently using same OS which I was using for Raspberry Pi 3 ?

I followed below steps to build ARM plugins but getting error -

1. sudo apt-get update
   sudo apt-get install -y git cmake scons build-essential

2. git clone --recurse-submodules --single-branch --branch=master https://github.com/openvinotoolkit/openvino_contrib.git

3. cd openvino_contrib/modules/arm_plugin

4. mkdir build && cd build

5. cmake -DInferenceEngineDeveloperPackage_DIR=opt/intel -DCMAKE_BUILD_TYPE=Release .. && make ON this command I faced below error -

Many Thanks...!!!!

Cheers Mira

[ilya-lavrenov]

@MiraGaikwad you need to build both OpenVINO main repo and OpenVINO contrib. You are trying to point to /opt/intel as InferenceEngineDeveloperPackage_DIR, but there is no such file named InferenceEngineDeveloperPackageConfig.cmake. Such file is created only when you build OpenVINO from scratch and it's located in the <build directory root>:

1. Build OpenVINO
2. Run cmake on OpenVINO Contrib with InferenceEngineDeveloperPackage_DIR pointing to OpenVINO build tree

Or using a single step cmake -DIE_EXTRA_MODULE=<openvino_contrib source dir>/modules/arm_plugin <openvino source dir> when OpenVINO is built with extra modules like ARM plugin (see https://github.com/openvinotoolkit/openvino_contrib/wiki#how-to-build-openvino-with-extra-modules)

[hbalasu1]

Hi @MiraGaikwad are you able to build based on @ilya-lavrenov suggestion?

[alex-ht]

I follow the wiki and it works:

git clone --recurse-submodules --single-branch --branch=releases/2022/1 https://github.com/openvinotoolkit/openvino_contrib.git
cd openvino_contrib/modules/arm_plugin
docker image build -t arm-plugin -f dockerfiles/Dockerfile.RPi64_focal .
mkdir build
docker container run --rm -ti -v $PWD/build:/arm_cpu_plugin arm-plugin

You will get OV_ARM_package.tar.gz under build/ folder.

[alex-ht]

@brmarkus - Thanks for the response. I checked the link and looks like those plugins are for Raspberry Pi4 but I am using Raspberry pi 3.

Do you know if plugins are available for Raspberry Pi3 as well ?

Cheers Mira

I'm using asus tinkerboard and I believe all ARM boards will work.

[MiraGaikwad]

@hbalasu1 - No it did not work for me yet. I am currently trying on Raspberry pi 3 but no luck.

[hbalasu1]

Hi @MiraGaikwad After further investigation, OpenVINO for arm CPU plugin only supports the following devices (based on openvino_contrib)

Raspberry Pi* 4 Model B	Debian* 9 (32-bit)
Raspberry Pi* 4 Model B	Debian* 10.3 (32-bit)
Raspberry Pi* 4 Model B	Ubuntu* 18.04 (64-bit)
Raspberry Pi* 4 Model B	Ubuntu* 20.04 (64-bit)
Apple* Mac mini with M1	macOS 11.1 (64-bit)
Samsung* Galaxy S20 FE	Android 10 (64-bit)

Unfortunately, OpenVINO does not support the Raspberry Pi 3 arm CPU, so the only option is only NCS2 if Raspberry Pi 3 is used.

Hope this information help

Thank you