WQ

`CUDA_VISIBLE_DEVICES=6，7 python ./mpi_from_images.py \ --image1=examples/twain/twain_left.jpg \ --image2=examples/twain/twain_right.jpg \ --output_dir=examples/twain/results \ --fx=1.722207943 \ --fy=2.296277257 \ --xoffset 0.0082 \ --render_multiples -2,-1.5,-1,-0.5,0,0.5 \ --render` There was a problem with the above input: `File...

Sorry to bother you, but I have a question. I have tried to use the MFIR test code to evalute the psnr of BIPNET about grayscale denoising. However, I can...

Hi z-bingo, I'm sorry to bother you with a question. I get a wrong points less than 20 dB and result with the gain of 8 while the result is...

关于HIN的params和flops？

1

因为HIN block这个块是自带卷积的，所以在测试中发现HIN带来的params和flops的增长是比较大的，所以想问一下paper里面的HIN消融实验的结果，params和flops是否进行过对齐？

We directly submitted the test on DND using your pre-trained model like other methods, but only got 38.42. Does this mean that NAFNET is prone to overfitting?

Dear author, According to the conclusion of your experiment based on the data proportion, fine-tuning only with the original SIDD data actually performs the best on the SIDD validation set....