Giter VIP home page Giter VIP logo

nssadnn_iqa's Introduction

NSSADNN_IQA

Pytorch version of IEEE Transactions on Multimedia 2019 : B. Yan, B. Bare and W. Tan, "Naturalness-Aware Deep No-Reference Image Quality Assessment," in IEEE Transactions on Multimedia, vol. 21, no. 10, pp. 2603-2615, Oct. 2019, doi: 10.1109/TMM.2019.2904879.

Note

*I did not use the learning rate that used in the paper: 0.01, because the ideal result cannot be obtained when the initial learning rate is 0.01, so the initial learning rate set here is 0.001 instead of 0.01. *This training progress only support on LIVE II database now, the training progress on TID2013, CSIQ, LIVEMD, CLIVE will be released soon.

Train

python train.py

TODO

  • Cross dataset test code will be published
  • Train on different distortion types on LIVE, TID2013, CSIQ will be published
  • Code of evaluations on Waterloo Exploration Database (D-test, L-test, P-test and gMAd competition) will be published

nssadnn_iqa's People

Contributors

lllllllllllll-llll avatar

Stargazers

avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar avatar

Watchers

avatar

Forkers

zhangyingyue jyothi6799 abandonsea yolandamdavis waihoh

nssadnn_iqa's Issues

How to test?

Thank you for your contribution. I trained with your code to get the model.pth file. How should I test it

The ref_name is not used?

@lllllllllllll-llll Thanks for your reply. I find ref_names in IQADataset.py is not used. What is the meaning of ref_names？And could you share the model with me? I want to conpare it to my own model. Thank you very much!!

The input of brisque function

Hi, thanks for your amazing work! I'm curious why you use cv2.BGR2RGB before inputing imges to the brisque function. Should the input be RGB images? Looking forward to your reply. thx~

NSSADNN_IQA/IQADataset.py

Line 108 in 1cc3774

im_features = cv.cvtColor(im_features, cv.COLOR_BGR2RGB)

KROCC and RMSE

Thank you for sharing. In the process of reproduction, I find that SRCC and PLCC are not very different from the original paper, but KROCC and RMSE are quite different from the original paper. I would like to ask why. Looking forward to your reply.

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.