zhuhancheng / MetaIQA

MetaIQA

• PyTorch 0.4.1 (with Python 3.6.0) implementation of the following paper.

• If you find our work is useful, pleaes cite our paper:
  @InProceedings{Zhu2020MetaIQA,
  author = {Zhu, Hancheng and Li, Leida and Wu, Jinjian and Dong, Weisheng and Shi, Guangming},
  title = {{MetaIQA:} Deep Meta-Learning for No-Reference Image Quality Assessment},
  booktitle = {Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR)},
  month = {Jun.},
  pages ={14143--14152},
  year = {2020}
  }

• MetaIQA_On_TID2013_KADID.py for quality prior model training.

• MetaIQA_FineTune_WILDLIVE.py for model fine-tuning on LIVE challenge database.

• If you want to train the quality prior model, you need download TID2013 and KADID-10K databases.