SIFID score for 8-bit grayscale images

[Kavit212]

Hi, Thank you very much for your codes. I have a small problem. All my images are grayscale 8-bit images. When I try to run the SIFID score on them, I get this error message.

Traceback (most recent call last): File "SIFID/sifid_score.py", line 258, in sifid_values = calculate_sifid_given_paths(path1,path2,1,args.gpu!='',64,suffix) File "SIFID/sifid_score.py", line 240, in calculate_sifid_given_paths m1, s1 = calculate_activation_statistics([files1[i]], model, batch_size, dims, cuda) File "SIFID/sifid_score.py", line 202, in calculate_activation_statistics act = get_activations(files, model, batch_size, dims, cuda, verbose) File "SIFID/sifid_score.py", line 100, in get_activations images = images[:,:,:,0:3] IndexError: too many indices for array

Could you kindly advice me on what changes to be made so that I can calculate the SIFID score for my images. Thank you.

[15732031137]

I also encountered this problem when running SR.py, this error occurs when a picture is 8-bit depth

[sw5park]

@Kavit212, @15732031137

Hello, were you able to solve this issue? I too am running training on grayscale images, and encounter the same problem