search

SSL92 / hyperIQA

Source code for the CVPR'20 paper "Blindly Assess Image Quality in the Wild Guided by A Self-Adaptive Hyper Network"
MIT License
346 stars 52 forks source link

Question about the performance of SOTA method. #35

Open xiezhangxiang opened 2 years ago

xiezhangxiang commented 2 years ago

hello, thank you for sharing your great work. I have some question about the performance of WaDIQaM, as shown in Table1, it achieved 0.955 SRCC and 0.973 PLCC on CSIQ database, how did you get the score? Was it trained on LIVE database?

lllllllllllll-llll commented 2 years ago

I also have the same question. I have retrained and tested WaDIQaM on CSIQ, but it hardly to achieve 0.910 SROCC.

lllllllllllll-llll commented 2 years ago

In addition, as we know, for CSIQ dataset, it's really hard to achieve 0.940 SROCC for ten dataset split process. I really suspect the results of WaDIQaM.