英文:
OpenVINO library does not see Neural Compute stick 2 on Raspberry pi 4
问题
I used this guide to install the openVINO toolkit on my Raspberry pi 4 model B. And this to perform inference on Neural Compute Stick 2.
But then when I tried to look at the available devices using the openvino library:
import openvino.inference_engine as ie
engine = ie.IECore()
print(engine.available_devices)
It shows:
['CPU']
But it also had to output 'MYRIAD'
In lsusb
i see my device:
Bus 001 Device 003: ID 03e7:2485 Intel Movidius MyriadX
How can I fix this?
I tried reinstalling other versions of Raspbian OS 64 bit but it didn't help
英文:
I used this guide to install the openVINO toolkit on my Raspberry pi 4 model B. And this to perform inference on Neural Compute Stick 2.
But then when I tried to look at the available devices using the openvino library:
import openvino.inference_engine as ie
engine = ie.IECore()
print(engine.available_devices)
It shows:
['CPU']
But it also had to output 'MYRIAD'
In lsusb
i see my device:
Bus 001 Device 003: ID 03e7:2485 Intel Movidius MyriadX
How can I fix this?
I tried reinstalling other versions of Raspbian OS 64 bit but it didn't help
答案1
得分: 1
1 中的Debian软件包列表不包括用于在Neural Compute Stick 2上进行推断所需的Myriad插件(libopenvino_intel_myriad_plugin.so),它们仅包括ARM CPU的CPU插件。您可以通过列出path-to/l_openvino_toolkit_debian9_2022.3.0.9052.9752fafe8eb_arm64/runtime/lib/aarch64目录来验证这一点,并查看libopenvino_intel_myriad_plugin.so是否丢失。
<br>
在Raspbian上获取Myriad插件的替代方法之一是使用2022.1分支从源代码构建OpenVINO。您可以按照文档中概述的说明来实现这一点。以下是一些帮助您进行此替代操作的资源:
英文:
The Debian packages listed in storage.intel.com do not include Myriad plugin (libopenvino_intel_myriad_plugin.so) needed to inference on Neural Compute Stick 2, they only include the CPU plugin for ARM CPU. You can verify this by listing the path-to/l_openvino_toolkit_debian9_2022.3.0.9052.9752fafe8eb_arm64/runtime/lib/aarch64 directory and see libopenvino_intel_myriad_plugin.so missing:
<br>
One of the alternatives to get the Myriad plugin on Raspbian is to build OpenVINO from source using 2022.1 branch. You can achieve this by following the instructions outlined in documentation. Here are some resources to help with this alternative:
通过集体智慧和协作来改善编程学习和解决问题的方式。致力于成为全球开发者共同参与的知识库,让每个人都能够通过互相帮助和分享经验来进步。
评论