Skip to content
This repository has been archived by the owner on Jan 25, 2023. It is now read-only.

Demo on iGPU for FFmpeg decode and scale, OpenVINO inference. this is zero-copy solution, which means No frame data copy from CPU to iGPU.

License

Notifications You must be signed in to change notification settings

intel-iot-devkit/ffmpeg-VAAPI-OpenVINO

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

15 Commits
 
 
 
 
 
 
 
 

Repository files navigation

DISCONTINUATION OF PROJECT

This project will no longer be maintained by Intel. Intel has ceased development and contributions including, but not limited to, maintenance, bug fixes, new releases, or updates, to this project. Intel no longer accepts patches to this project.

ffmpeg-VAAPI-OpenVINO

Sample for FFmpeg with VAAPI hardware decode and scale, combining OpenVINO iGPU inference

For FFmpeg user/customer, we provide a solution about FFMPEG iGPU hardware decode connectiong OpenVINO iGPU inference engine avoiding frame data copy. The frame data after decoding/scaling in the buffer can be directly read by iGPU’s inference engine, so that we can have low CPU utilization and maxinum iGPU usage.
The whole workload runs on iGPU, compared with before ,we have fast pre-process speed and more bandwidth between CPU/iGPU communication.

Dependences:

  • Intel(R) OpenVINO(TM) Toolkit
  • FFmpeg
  • VA-API (libVA)

Device request:

Intel(R) CPU with Integrated Graphics

Validated software version:

  • OpenVINO 2020.3 LTS
  • FFmpeg 4.3.1
  • LibVA 2.1.0, VA-VAPI 1.6.0 Intel iHD driver 19.3.1

About

Demo on iGPU for FFmpeg decode and scale, OpenVINO inference. this is zero-copy solution, which means No frame data copy from CPU to iGPU.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages