This is a demonstration of the Combining Tensor Slice and Singular Value for Blind Light Field Image Quality Assessment(TSSV-LFIQA)described in:
Zhiyong Pan, Mei Yu, Gangyi Jiang, Haiyong Xu, and Yo-Sung Ho, " Combining Tensor Slice and Singular Value for Blind Light Field Image Quality Assessment," IEEE Journal of Selected Topics in Signal Processing, 15(3): 672-687, 2021, DOI: 10.1109/JSTSP.2021.3056959.
% Code provided by Zhiyong Pan 2020.10.23
The code was written in MATLAB 2017a, and tested on Windows 10/7.
An exmaple ( Matlab codes, light field image and mos in Win5-LID[1]) are provided to demonstrate how to use the package.
I. RUNNING CODE
- Feature Extration and Quality Prediction
run Demo_TSSV_LFIQA.m
- Regression
run Demo_SVR.m
II. Package Composition
- Feature
(1)Tensor Slice Spatial Feature (Including First_Grad_Feature.mat;First_Grad_Feature_Color.mat;Other_Entropy_Feature.mat) (2)Singular Value Angular Feature (Including SinValue_inter_Feature.mat;SinValue_intra_Feature.mat)
-
Functions (Subfunction)
-
Libraries（Including SVR and tensor_toolbox）
-
Demo_SVR.m (Regression model)
-
Demo_TSSV_LFIQA.m (Main Function)
-
LN_dishes_50.bmp （Light field image in Win5-LID[1]）,downloaded in "https://pan.baidu.com/s/1_CNTo2kJxP1UMJxpuk7XRA" password：4xr9
-
model.mat (Model trained on Win5-LID[1])
-
win5mos.mat (MOS in Win5-LID[1],sorted by scene)
-
Readme.txt(Manual)
[1]L. Shi, S. Zhao, W. Zhou, and Z. Chen, "Perceptual evaluation of light field image," in Proc. IEEE Int. Conf. Imag. Process., Athens, Greece, 2018, pp. 41-45.
III. VERSION HISTORY
v1.0 - Initial release