Integrating Remote tensor with directx context in openvino

[Gayathri_Sankaran]

Hi Team,

I am currently working on a project where I need to integrate oepnvino inference using remote tensor context for the object detection model. I am using windows system, the rendering pipeline exist in directx context. I need to pass the image frame information in the gpu to the remote tensor context and need to compile the model in remote device and do the inference for the same. The entire pipeline  should be happening in the GPU. The documentation for using directx implementation of this type is very limited. Can you provide some additional examples or documentation for this? 

Specifically, any best practices, common pitfalls and complete example code would be highly appreciated.

Thank you for your assistance !

[Iffa_Intel]

Hi,

perhaps this documentation would help your use case (refer to Running GPU Plugin Inference within user-supplied shared context section)  

Cordially,

Iffa

[Gayathri_Sankaran]

Thanks for the documentation. Any sample examples for the same available from the Intel community in git repo there?

[Iffa_Intel]

Hi,

we'll clarify for other sample example resources (if available) and get back to you asap.

Cordially,

Iffa

[Iffa_Intel]

Hi,

After some clarification, there are no other guides/samples than the previously shared link as of now.

Cordially,

Iffa

[Iffa_Intel]

Hi,

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

Cordially,

Iffa