Intel® Distribution of OpenVINO™ Toolkit
Community assistance about the Intel® Distribution of OpenVINO™ toolkit, OpenCV, and all aspects of computer vision-related on Intel® platforms.

reuse of inferRequest

ben__ben
Beginner
651 Views

Hello folks!

I'm developing an application in c, and using the inference engine provided by the openvino.

I found that, the inference process is very fast, but the bottleneck is creating the inferRequest. It took about 80ms on my computer.

ExecutableNetwork executable_network = plugin.LoadNetwork(network, {});
InferRequest infer_request = executable_network.CreateInferRequest();

since my application is in c. the application calls the c++ API(dll) many times. The overhead is too huge.

is it possible to new the infer_request such that I don't have to create the inferRequest everything I call the c++ interface ?

thanks!

 

0 Kudos
1 Reply
Shubha_R_Intel
Employee
651 Views

Dear ben, ben,

Please consider these Inference Engine performance topics (such as Async API and Throughput Mode), mentioned in the below document:

http://docs.openvinotoolkit.org/latest/_docs_IE_DG_Intro_to_Performance.html

Thanks,

Shubha

0 Kudos
Reply