Inference Engine Python API documentation

[Cariaggi__Francesco]

Hello everyone,

I've succesfully installed OpenVINO 2019 R1 on my Linux machine, so now I would like to take advantage of the Inference Engine to develop some Python applications. However, documentation for the Python API is nowhere to be found. I was able to find some Python examples within the installation directory and, by looking at the source code, I found a line that looked like this:

from openvino.inference_engine import IENetwork, IEPlugin

so I assume that the most important classes to use are IENetwork (to load models in IR format) and IEPlugin (to decide where to run inference, whether it is on the CPU, GPU, Myriad X VPU or FPGA). But again, I found no documentation at all. Can anyone help me?

[Mikhail_T_Intel · Accepted solution]

Hi Francesco,

The Python API reference can be found here.

Regarding the classes, you are right the only two classes which you need to import and create instances manually its IENetwork and IEPlugin, all the other mentioned in documentation classes will be created using those two classes functions, e.g. IEExecutableNetwork will be created after loading of IENetwork instance to the plugin.

[WDomo]

Supported OSes

Currently the Inference Engine Python* API is supported on Ubuntu* 16.04 and 18.04, Windows* 10, macOS* 10.x and CentOS* 7.3 OSes. Supported Python* versions:

When are we likely to see Raspbian supporting the API?

[Shubha_R_Intel]

Dear Domoney, William,

Python API has always been supported in Raspbian. Looks like this info is missing from the docs though.

Please find the R3 version of Raspbian OpenVino here:

https://download.01.org/opencv/2019/openvinotoolkit/R3/

Thanks,

Shubha