Hi Radu,
When I submitted the results for the case of x4, I always got the following error
Traceback (most recent call last):
File "/codalabtemp/tmpde45KX/run/program/evaluation.py", line 100, in
compute_psnr(ref_im,res_im)
File "/codalabtemp/tmpde45KX/run/program/evaluation.py", line 38, in compute_psnr
_open_img(os.path.join(input_dir,'res',res_im))
File "/codalabtemp/tmpde45KX/run/program/evaluation.py", line 17, in output_psnr_mse
squared_error = np.square(img_orig - img_out)
ValueError: operands could not be broadcast together with shapes (1180,2020,3) (1516,2020,3)
It seems it is a problem of the image size. However, I didn't get any error when I submitted the results for the cases of x2 and x3. Moreover, I have checked the dimension of recovered x4 images and the dimension of recovered x3 images, one image by one image, but I didn't find any difference. I'm very confused where this error comes from?
Best regards,
Yunjin
Posted by: chenyunjin_ULSee @ April 13, 2017, 4:36 a.m.Thanks Yunjin for pointing out this problem both here and by email.
Meanwhile this problem had been fixed.
Success,
Radu