Code Monkey home page Code Monkey logo

censeoqoe's Introduction

CenseoQoE: 视觉感知画质评价框架

PRs Welcome

项目介绍

图像/视频在编解码、传输和显示等过程中难免引入不同类型/程度的失真导致图像质量下降。图像/视频质量评价(IVQA)的研究目标是希望模仿人类视觉感知系统, 通过算法评估图片/视频在终端用户的眼中画质主观体验的好坏,目前在视频编解码、画质增强、画质监控、推荐算法、竞品分析等领域有广泛的应用。

CenseoQoE 提供图像/视频画质评价算法模型训练到应用落地的完整方案,主要包括 CenseoQoE-AlgorithmCenseoQoE-SDK两部分。 CenseoQoE-Algorithm 是通用的画质评价算法模型的训练框架,CenseoQoE-SDK集成了针对多种业务场景训练好的画质评价模型,可直接对UGC视频、PGC视频、游戏视频等做无参考/有参考的画质评价,实现开箱即用。

为什么叫CenseoQoECenseo 在拉丁语中是我认为/我觉得的意思,与图像/视频画质体验评价领域的主观评测概念相符,QoE则是体验质量(Quality of Experience)。

News

Our two papers based on CenseoQoE have been accepted by ICASSP 2022

Please refer to these two papers for more detailed information. More released pretrained models can be found at CenseoQoE-Algorithm.

If our code helps your research, please consider citing these two papers.

主要特性

  • CenseoQoE-Algorithm 提供图像/视频画质评价算法模型训练的codebase,集成自研的模型以及业界开源的算法模型,支持无参考和有参考。此外,我们会公开不同业务场景下 (UGC视频、PGC视频、游戏视频、直播视频等)的预训练模型,用户可以利用这些预训练模型在自己的数据集进行微调优化。
  • CenseoQoE-SDK 集成了基于CenseoQoE-Algorithm在我们自研的数据集上训练得到的多个业务场景下的画质评价模型,全部使用c++实现,编译之后可直接 对UGC视频、PGC视频、游戏视频等做画质评价,实现开箱即用。

快速开始

  • 如果你想做模型的训练或是优化改进,请参见CenseoQoE-Algorithm
  • 如果你想快速使用CenseoQoE推出的画质评价工具/SDK,请参见CenseoQoE-SDK

Demo

以下是我们针对游戏视频训练的两个模型(resnet-based和shufflenet-based)和vmaf的两个模型(vmaf-default和vmaf-phone)在同一个视频上的质量分数对比。

our model vs vmaf

常见问题

请参考 FAQ 了解其他用户的常见问题。

许可

该项目使用的开源协议为 MIT

引用

@misc{wen2021strong,
      title={A strong baseline for image and video quality assessment}, 
      author={Shaoguo Wen and Junle Wang},
      year={2021},
      eprint={2111.07104},
      archivePrefix={arXiv},
      primaryClass={eess.IV}
}

参与贡献

我们非常欢迎用户对于 CenseoQoE 做出的任何贡献,可以参考 CONTRIBUTION.md 文件了解更多细节。

致谢

我们感谢所有为该项目提供建议、数据、模型、使用样例和新功能支持的贡献者,以及提供宝贵反馈的用户。

censeoqoe's People

Contributors

shaoguowen avatar

Stargazers

 avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar

Watchers

 avatar  avatar  avatar  avatar  avatar  avatar  avatar

censeoqoe's Issues

SDK执行无输出结果

请问本地编译之后,执行,输出是这样的对吗,但是resulsts目录下面没有内容,怎么看分析的结果呢
image

分数出现负数

在使用train_code中inference.py使用UGC视频对应的pytorch预训练模型对项目中test_img进行无参照评分时出现负数,以下是输出结果
{"test.jpg": -3.90386700630188}

English

Hi, It would help me if this page is in English language.
Sorry,

Thanks.

请教一个关于视频抽帧的问题

每个视频因为运动信息强弱,不同帧可能会有不同的质量,如果像Konvid这种单个打分的视频数据集,对视频帧都赋予同一个打分是不是不合理?请教下这个问题是否有解决办法。

Is DMOS value necessary?

Hi shaoguo, thanks for ur great job!!!
Now, i have the MOS value of images but not DMOS, could i use this repo for IQA trainning? I wanna know if the DMOS value is necessary in trainning model. Looking forward to ur reply and thanks in advance!!!

Recommend Projects

  • React photo React

    A declarative, efficient, and flexible JavaScript library for building user interfaces.

  • Vue.js photo Vue.js

    🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.

  • Typescript photo Typescript

    TypeScript is a superset of JavaScript that compiles to clean JavaScript output.

  • TensorFlow photo TensorFlow

    An Open Source Machine Learning Framework for Everyone

  • Django photo Django

    The Web framework for perfectionists with deadlines.

  • D3 photo D3

    Bring data to life with SVG, Canvas and HTML. 📊📈🎉

Recommend Topics

  • javascript

    JavaScript (JS) is a lightweight interpreted programming language with first-class functions.

  • web

    Some thing interesting about web. New door for the world.

  • server

    A server is a program made to process requests and deliver data to clients.

  • Machine learning

    Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.

  • Game

    Some thing interesting about game, make everyone happy.

Recommend Org

  • Facebook photo Facebook

    We are working to build community through open source technology. NB: members must have two-factor auth.

  • Microsoft photo Microsoft

    Open source projects and samples from Microsoft.

  • Google photo Google

    Google ❤️ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.