Releases: intel/libvpl
v2021.2.2
This release is focused on alignment with the 2.2 oneVPL specification, enabling GPU support, and delivering feature support for partners and customers.
The Intel® oneAPI Video Processing Library (oneVPL) lets developers add high- speed, real-time decoding, encoding, and video processes to their applications. Its single video API provides direct access to advanced Intel® CPU instructions and gives you total control of video processing needs.
This release delivers the following new features:
• Dispatcher and CPU implementation updated to align with oneVPL Specification 2.2
• Legacy Intel(R) Media SDK test tools (with CPU and GPU support)
• Intel(R) Media SDK to oneVPL migration guide
• pkg-config support
• Windows* 32-bit support
• Legacy mfxvideo++ header
• New Preview Sample:
-- oneVPL-OpenVINO™ Interop Sample (Preview) (source code only)
• New OS support:
-- Ubuntu 20.10
v2021.1
Intel® oneAPI Video Processing Library (oneVPL) initial release
The Intel® oneAPI Video Processing Library (oneVPL) lets developers add high-speed, real-time decoding, encoding, and video processes to their applications. Its single video API provides direct access to advanced Intel® CPU instructions and gives you total control of video processing needs.
oneVPL delivers the following features:
• C API implementation of the oneVPL API 2.0
• oneVPL Dispatcher
• New frame memory interface with access to internally allocated buffers
• CPU implementation including
- Software encode and decode support for :
-- H.264/AVC
-- H.265/HEVC
- MJPEG
-- AV1
- DPC++ kernel integration
- Sample support with:
-- hello-decode
-- hello-encode
-- hello-vpp
-- dpcpp-blur
-- hello-transcode (open source only)
- The base toolkit package includes pre-built tools (vpl-decode, vpl-encode, vpl-vpp)