img_dim = L_high_np.shape[1:]
sample(sample_imgs,
split=split_point,
figure_size=(2, 3),
img_dim=img_dim,
path=filepath,
num=epoch,
metrics=True)
if __name__ == "__main__":
criterion = Restore_Loss()
model = RestoreNet_Unet(use_MaskMul=False)
decom_net = DecomNet()
parser = BaseParser()
args = parser.parse()
with open(args.config) as f:
config = yaml.load(f)
args.checkpoint = True
if args.checkpoint is not None:
decom_net = load_weights(decom_net,
path='./weights/decom_net_normal.pth')
log('DecomNet loaded from decom_net.pth')
model = load_weights(model, path='./weights/restore_net_finetune.pth'
) # restore-SID/restore_mask_0.pth')
log('Model loaded from restore_net.pth')
root_path_train = r'H:\datasets\Low-Light Dataset\KinD++\LOLdataset\our485'
from base_parser import BaseParser
from memory_parser import MemoryParser
from file_list import filename_list
base = BaseParser()
# base.read_cosrad_table("_cosrad/82_2020-3-3_23i16i27.xls")
BRIEF_DATA = "brief_data.txt"
def create_dir(path_dir):
import os
if os.path.isdir(path_dir) is False:
os.mkdir(path_dir)
memory = MemoryParser(BRIEF_DATA, False)
for file, cosrad in filename_list:
print("{0:s} is being processed".format(file))
memory.error_parse(file, cosrad)
