def test(args, saved_model_path, noise, famous_path, testset_path=None):
"""Run predictable test
"""
torch.manual_seed(7)
model = restore_model(args, saved_model_path)
if USE_CUDA:
model = model.cuda()
norm_noise = common.normilize(noise, 255)
padding = 20
if testset_path is not None and os.path.isdir(testset_path):
testset = create_test_dataset(testset_path, noise, padding)
test_loader = DataLoader(testset)
ours_psnr, bm3d_psnr = avarge_psnr_testset(model, test_loader,
padding, norm_noise)
else:
print('testset path was not provided or does not exsist on machine'
+' skipping to famouse images testset')
ours_psnr = bm3d_psnr = 0
testset = create_famous_dataset(famous_path, noise, padding)
file_names = testset.image_filenames
famous_loader = DataLoader(testset)
fam_psnrs, fam_res_array =\
famous_images_teset(
model,
famous_loader,
file_names,
padding,
norm_noise)
return fam_psnrs, fam_res_array, file_names, ours_psnr, bm3d_psnr
def input_process_fn(_x):
return gaussian(_x,
is_training=True,
mean=0,
stddev=normilize(noise, 255))
def pre_process_fn(_x):
return normilize(nhwc_to_nchw(_x), 255)
def pre_process_fn(_x):
return normilize(_x, 255)