open_model_zoo
open_model_zoo copied to clipboard
Using oneVPL and QSV/Intel GPU decoder in images_capture
Any plans on adding support for oneVPL backend so that open_images_capture() can take advantage of hardware decoder in iGPU (& upcoming dGPU)?
There are no such plans
Is there any demo that uses QSV? oneVPL backend?
No
@js333031 there is demo based on G-API, which enables oneVPL, but it is under review: https://github.com/openvinotoolkit/open_model_zoo/pull/3535
Hi @js333031
The plan was to enable the GPU acceleration for decode, pre- and post-processing in OpenCV G-API and so bring it to the OMZ pipelines semi-automatically. If it works for you, we can prioritize it when possible.
Dmitry
Ideally, it would be to have the GPU acceleration for decode, pre- and post-processing in simplest possible demos. I assume those are the OMZ python demos. Can this be added to OMZ python demos?