Add GTLQ dataset

This commit is contained in:
James Betker 2020-04-22 00:40:38 -06:00
parent 4d269fdac6
commit 12d92dc443
2 changed files with 7 additions and 5 deletions

View File

@ -7,14 +7,14 @@ import torch.utils.data as data
import data.util as util
class LQGTDataset(data.Dataset):
class GTLQDataset(data.Dataset):
"""
Read LQ (Low Quality, e.g. LR (Low Resolution), blurry, etc) and GT image pairs.
If only GT images are provided, generate LQ images on-the-fly.
Reads unpaired high-resolution and low resolution images. Downsampled, LR images matching the provided high res
images are produced and fed to the downstream model, which can be used in a pixel loss.
"""
def __init__(self, opt):
super(LQGTDataset, self).__init__()
super(GTLQDataset, self).__init__()
self.opt = opt
self.data_type = self.opt['data_type']
self.paths_LQ, self.paths_GT = None, None
@ -43,7 +43,7 @@ class LQGTDataset(data.Dataset):
self._init_lmdb()
GT_path, LQ_path = None, None
scale = self.opt['scale']
GT_size = self.opt['GT_size']
GT_size = self.opt['target_size']
# get GT image
GT_path = self.paths_GT[index]

View File

@ -32,6 +32,8 @@ def create_dataset(dataset_opt):
from data.LQ_dataset import LQDataset as D
elif mode == 'LQGT':
from data.LQGT_dataset import LQGTDataset as D
elif mode == 'GTLQ':
from data.GTLQ_dataset import GTLQDataset as D
# datasets for video restoration
elif mode == 'REDS':
from data.REDS_dataset import REDSDataset as D