Official PyTorch code for the SPL2025 paper "Blind Light Field Image Quality Assessment via Frequency Domain Analysis and Auxiliary Learning". Please refer to the paper for details.

Note: First, we convert the dataset into H5 files using MATLAB. Then, we train and test the model in Python.

Generate Dataset in MATLAB

Take the NBU-LF1.0 dataset for instance, convert the dataset into h5 files, and then put them into './Datasets/NBU_FABLFQA_5x64x64/':

 ./FABLFQA/Datasets/Generateh5_for_NBU_Dataset.m

Train

Train the model using the following command:

python Train.py  --trainset_dir ./Datasets/NBU_FABLFQA_5x64x64/

Test Overall Performance

Test the overall performance using the following command:

python Test.py

Test Individual Distortion Type Performance

Test the individual distortion type performance using the following command:

 python Test_Dist.py

Acknowledgement

This project is based on DeeBLiF. Thanks for the awesome work.

Citation

Please cite the following paper if you use this repository in your reseach.

@ARTICLE{10844526,
  author={Zhou, Rui and Jiang, Gangyi and Zhu, Linwei and Cui, Yueli and Luo, Ting},
  journal={IEEE Signal Processing Letters}, 
  title={Blind Light Field Image Quality Assessment via Frequency Domain Analysis and Auxiliary Learning}, 
  year={2025},
  volume={},
  number={},
  pages={1-5},
  keywords={Measurement;Feature extraction;Discrete cosine transforms;Distortion;Light fields;Three-dimensional displays;Spatial resolution;Visualization;Frequency conversion;Frequency-domain analysis;Light field;blind image quality assessment;frequency domain;auxiliary learning;deep learning network},
  doi={10.1109/LSP.2025.3531209}}

Contact

For any questions, feel free to contact: [email protected] or [email protected]