jizhongpeng / uavs3e

AVS3 encoder which supports AVS3-P2 baseline profile.

uavs3e

uavs3e is an opensource avs3 encoder, supports AVS3-P2 baseline profile.
features:

1. AVS3-P2 baseline profile.
2. supports to compile for Windows/Linux systems.
3. optimized for SSE4/AVX2 chips.
4. 10bit encoding on all supported platforms (without SIMD).
5. The uavs3 codec has supported x86 and arm platforms, and has been tested and verified on the Kunpeng processor.
6. The ARM platform recommends the Kunpeng processor.

license

Copyright reserved by “Peking University Shenzhen Graduate School”, “Peng Cheng Laboratory”, and “Guangdong Bohua UHD Innovation Corporation”

This program is a free software. You can redistribute it and/or modify it under the terms of the BSD 4-clause license.
For more details, please view the file "COPYING" in the project.

compile

The default configuration only support 8bit decoding.
To support 10bit streams decoding:cmake -DCOMPILE_10BIT=1

windows

Prerequisites: Visual Studio 2017

build:

1. ./version.bat (to generate version.h)
2. solution file: build/x86_windows/uavs3e.sln

To support 10bit streams decoding, edit inc/com_api.h : #define COMPILE_10BIT 1

linux

Prerequisites:

1. gawk (http://www.gnu.org/software/gawk/)
2. CMake (https://cmake.org) version 2.8 or higher

Build:

1. mkdir build/linux
2. cd build/linux && cmake -DCOMPILE_10BIT=0 ../..
3. make && make install

To support 10bit streams decoding:cmake -DCOMPILE_10BIT=1 to build shared library, set BUILD_SHARED_LIBS=1 please.

Run tests

window/linux

uavs3e [--config file] [-paramShort ParameterValue] [--paramLong ParameterValue]

 --config file    
         All Parameters are initially taken from the 'file', typically: "encode_RA.cfg".

 -paramShort ParameterValue
 --paramLong ParameterValue
         If -paramShort or --paramLong parameters are present, then the ParameterValue will override the default settings in the configuration file.

Contributors

This program is originally developed by the team of Prof.Ronggang Wang ([email protected]) at Peking University Shenzhen Graduate School.

• Main contributors:

  • Zhenyu Wang ([email protected]), Peking University Shenzhen Graduate School.
  • Bingjie Han ([email protected]), Peking University Shenzhen Graduate School.
  • Jiang Du, Peking University Shenzhen Graduate School.
  • Kui Fan, Peking University Shenzhen Graduate School.
  • Xi Xie, Peking University Shenzhen Graduate School.
  • Shengyuan Wu, Peking University Shenzhen Graduate School.
  • Tong Wu, Peking University Shenzhen Graduate School.
  • Shiyi Liu, Peking University Shenzhen Graduate School.
  • Jiayu Yang, Peking University Shenzhen Graduate School.
  • Chunhui Yang, Peking University Shenzhen Graduate School.
  • Dapeng Chen, Bigo Technology Pte. Ltd
  • Jin Lin, Harbin Institute of Technology, Shenzhen.
  • Yi Ma, Peking University Shenzhen Graduate School.
  • Litian Li, Peking University Shenzhen Graduate School.
  • Guisen Xu, Peking University Shenzhen Graduate School.
  • Xufeng Li, Peking University Shenzhen Graduate School.
  • Yangang Cai, Peking University Shenzhen Graduate School.
  • Hao Lv, Peng Cheng Laboratory.
  • Ronggang Wang ([email protected]), Peking University Shenzhen Graduate School

  XueliCheng

  luofl

References

*This program also refers to the following fast algorithm:

• 陈焕浜, 杨海涛. SMVD编码加速. AVS技术提案, M5218. 2020年3月.
• 赵寅, 杨海涛. 块划分快速算法. AVS技术提案, M5137. 2020年3月.
• 张焕宸, 陈雅梅, 喻莉. 基于灰度共生矩阵与编码信息的帧内预测快速算法. AVS技术提案, M5327.