[Jiebin Yan], [Jiale Rao], [Xuelin Liu], [Yuming Fang], [Yifan Zuo], [Weide Liu]

You can be downloaded at the Baiduyun (Password: JUFE)

If you want to retrain the OIQAND model, using JUFE-10K database or another database, you first need to prepare viewport images.

• In get_viewport_images folder run demo.m

run demo.m

Then, you can select the corresponding training and test files under the file/JUFE-10K folder.

Modify the configuration in Original_code/config.py

• Modify "dataset_name" to choose which datasets you want to train in config
• Modify training and test dataset path

sh run.sh

If you want to directly test OIQAND model trained on JUFE-10K, you can download Weights at Baiduyun (Password: jufe)

sh test-run.sh

NOTE: The code in the code folder is the refactoring code of OIQAND model. I hope it will be easy for you to understand about OIQAND model

If you find this code is useful for your research, please cite:

@article{yan2024oiqand,
  title={Subjective and Objective Quality Assessment of Non-Uniformly Distorted Omnidirectional Images},
  author={Yan, Jiebin and Rao, Jiale and Liu, Xuelin and Fang, Yuming and Zuo, Yifan and Liu, Weide},
  journal={IEEE Transactions on Multimedia},
  year={2024}
}

1. https://github.com/LXLHXL123/JUFE-VRIQA
2. https://github.com/TianheWu/Assessor360