search

yuanjunchai / IKC

Implementation of 'Blind Super-Resolution With Iterative Kernel Correction' (CVPR2019)
Apache License 2.0
218 stars 44 forks source link

Testing for scale 2 #12

Open sooyekim opened 4 years ago

sooyekim commented 4 years ago

Hi,

Thanks for providing the code!

I tried testing for scale 2 by modifying the scale parameter in test_Predictor.yml and test_Corrector.yml, but the network keeps on upscaling the image by scale 4. Could you provide some more details for upscaling with scale factor 2?

Thanks!

msxiaojin commented 4 years ago

Try to modify the 'network_G': upscale =4 to upscale =4 in both .yml files. However, when I tested for scale 2, the output SR image seemed strange. Most of the pixel intensities are zero. Do you have the same problem?

sooyekim commented 4 years ago

@msxiaojin I tried modifying the scale parameter in test_SFTMD.yml to scale=2, which gave me x2 up-scaled results but with very dark pixel values like you. I suppose they haven't provided the pre-trained weights for scale factor 2, and the scale parameter in the .yml files is intended just for the training stage.