One thing I wanna inquire about is the execution procedure of the generate_rank_txt_live.py file.
It seems that images of four distortion types, i.e., JPEG, JP2K, Gblur, and GN, are required for this Python script.
After I downloaded the Waterloo Dataset and the corresponding toolbox, I found a distortion_generation function. Is there any script to generate the distorted images?
Thanks in advance and I'm looking forward to hearing from you.
I have found that following the codes here: https://github.com/xialeiliu/RankIQA/tree/master/data/data_generator can solve the problem and generate the dataset.
Dear Xialei,
Thanks a lot for your great work. It helps a lot.
One thing I wanna inquire about is the execution procedure of the
generate_rank_txt_live.py
file. It seems that images of four distortion types, i.e., JPEG, JP2K, Gblur, and GN, are required for this Python script. After I downloaded the Waterloo Dataset and the corresponding toolbox, I found a distortion_generation function. Is there any script to generate the distorted images?Thanks in advance and I'm looking forward to hearing from you.
Best regards,
Bruce Shuyue Jia