Some commonly used traditional image quality Assesment methods
install requeried tools: Matlab c++$ readme README.md
// Checks readme in current working directory
You need:
1)define your own data and train-test split
2)run Comp_main.m (only for methods: BRISQUE FRIQUEE GLBP) to get features
3)run SSEQ/SSEQ_MAIN.m for methods:SSEQ
4)run niqqe_release/example.m for methods:NIQE
notice: you need to change some para to run those methods because diff num of train-set
for example in IQA_predict.m, you should change the length of train_split and test_split
5)run tool_mean_std.m for calculating the mean and std performance