AHIQ
AHIQ copied to clipboard
[CVPRW 2022] Attentions Help CNNs See Better: Attention-based Hybrid Image Quality Assessment Network
Question for the results in the original paper: How is the evaluation on traditional datasets (LIVE, CSIQ, TID) performed? Do you report average performance over K runs? The paper only...
Hi, could you provide us regarding the PIPAL.txt and PIPAL_NTIRE_Valid_MOS.txt? I could not find any related files about that Thank you, I am looking forward for your reply.
Thank you for this great work and for sharing the code. Currently the challenge website does not provide the dataset for training and testing. Could you please share the data...
The test code is not working. can you please check and also provide sample input?
File "C:\Users\s4.gupta\Downloads\AHIQ-main\test.py", line 94, in load_model assert found, 'Model for epoch %i not found' % self.opt.load_epoch AssertionError: Model for epoch 0 not found I am confused with input text file...
Hi. Thanks for this interesting work. I want to use this full reference metric to evaluate the quality of some images for my research. I managed to run the code...
In paper section 4.2, the author says to use the AdamW optimizer. But actually in the code train.py, the author use the Adam. Does anyone know if the two optimizers...
The method uses the features of stage2 of Resnet50 as the result of feature extraction. It's very strange here, why choose stage2, or why not choose one output per stage...
I am trying to run your model, and have already setup the environments. I found this problem regarding vit_base_patch8_224 do you have any ideas to solve it? Thank you! really...
Hi, thanks for your hard work and congratulations for your 1st place in the CPVR competition, I am curious on the implementation of your model. and currently, I have tried...