CVIQDatabase

Introduction

CVIQ Database includes 524 compressed images generated from 16 original images. Three popular coding technologies are deployed in the database, which are JPEG, H.264/AVC, and H.265/HEVC, respectively.

Download

Baidu Netdisk:

Link:https://pan.baidu.com/s/1b3j5iO3jQWf2fZPAli9ozw  Password:b17u

Google Drive:

Link:https://drive.google.com/open?id=12E-sDZOq0DfCtNNwdyer7azfLZNNva6N

The images with names from 1*n to 11*n are compressed by JPEG, with names from 12*n to 22*n are compressed by AVC, with names from 23*n to 33*n are compressed by HEVC, and with names 34*n are original images.

The CVIQ.mat file contains the MOS for each image.

MC360IQA

The MC360IQA model can be download from: https://github.com/sunwei925/MC360IQA.

Citation

If you use CVIQ Database in your research, please cite:

@article{sun2019mc360iqa,
  title={MC360IQA: A Multi-channel CNN for Blind 360-degree Image Quality Assessment},
  author={Sun, Wei and Min, Xiongkuo and Zhai, Guangtao and Gu, Ke and Duan, Huiyu and Ma, Siwei},
  journal={IEEE Journal of Selected Topics in Signal Processing},
  volume={14},
  number={1},
  pages={64-77},
  year={2020},
  publisher={IEEE}
}

@inproceedings{sun2018large,
  title={A Large-Scale compressed 360-degree spherical image database: From subjective quality evaluation to objective model comparison},
  author={Sun, Wei and Gu, Ke and Ma, Siwei and Zhu, Wenhan and Liu, Ning and Zhai, Guangtao},
  booktitle={2018 IEEE 20th International Workshop on Multimedia Signal Processing (MMSP)},
  pages={1--6},
  year={2018},
  organization={IEEE}
}