DL-Art-School/codes/test.py

109 lines
3.8 KiB
Python
Raw Normal View History

import os.path as osp
import logging
2021-06-11 21:31:10 +00:00
import random
import time
import argparse
from collections import OrderedDict
import utils
import utils.options as option
import utils.util as util
2020-12-18 16:18:34 +00:00
from trainer.ExtensibleTrainer import ExtensibleTrainer
from data import create_dataset, create_dataloader
from tqdm import tqdm
import torch
2021-06-11 21:31:10 +00:00
import numpy as np
2021-06-11 21:31:10 +00:00
def forward_pass(model, data, output_dir, opt):
2021-04-22 00:14:17 +00:00
alteration_suffix = util.opt_get(opt, ['name'], '')
denorm_range = tuple(util.opt_get(opt, ['image_normalization_range'], [0, 1]))
2021-06-12 02:50:07 +00:00
with torch.no_grad():
model.feed_data(data, 0, need_GT=need_GT)
model.test()
visuals = model.get_current_visuals(need_GT)['rlt'].cpu()
2021-04-22 00:14:17 +00:00
visuals = (visuals - denorm_range[0]) / (denorm_range[1]-denorm_range[0])
fea_loss = 0
2020-10-27 16:25:42 +00:00
psnr_loss = 0
for i in range(visuals.shape[0]):
img_path = data['GT_path'][i] if need_GT else data['LQ_path'][i]
img_name = osp.splitext(osp.basename(img_path))[0]
sr_img = util.tensor2img(visuals[i]) # uint8
# save images
suffix = alteration_suffix
if suffix:
save_img_path = osp.join(output_dir, img_name + suffix + '.png')
else:
save_img_path = osp.join(output_dir, img_name + '.png')
if need_GT:
2020-10-27 16:25:42 +00:00
psnr_sr = util.tensor2img(visuals[i])
psnr_gt = util.tensor2img(data['hq'][i])
2020-10-27 16:25:42 +00:00
psnr_loss += util.calculate_psnr(psnr_sr, psnr_gt)
util.save_img(sr_img, save_img_path)
2020-10-27 16:25:42 +00:00
return fea_loss, psnr_loss
2020-07-24 18:26:44 +00:00
2020-04-24 05:59:09 +00:00
if __name__ == "__main__":
2021-06-11 21:31:10 +00:00
# Set seeds
torch.manual_seed(5555)
random.seed(5555)
np.random.seed(5555)
#### options
torch.backends.cudnn.benchmark = True
2020-11-29 19:21:31 +00:00
want_metrics = False
parser = argparse.ArgumentParser()
2021-06-14 15:14:30 +00:00
parser.add_argument('-opt', type=str, help='Path to options YAML file.', default='../options/test_diffusion_unet.yml')
opt = option.parse(parser.parse_args().opt, is_train=False)
opt = option.dict_to_nonedict(opt)
utils.util.loaded_options = opt
util.mkdirs(
(path for key, path in opt['path'].items()
if not key == 'experiments_root' and 'pretrain_model' not in key and 'resume' not in key))
util.setup_logger('base', opt['path']['log'], 'test_' + opt['name'], level=logging.INFO,
screen=True, tofile=True)
logger = logging.getLogger('base')
logger.info(option.dict2str(opt))
#### Create test dataset and dataloader
test_loaders = []
for phase, dataset_opt in sorted(opt['datasets'].items()):
test_set = create_dataset(dataset_opt)
test_loader = create_dataloader(test_set, dataset_opt)
logger.info('Number of test images in [{:s}]: {:d}'.format(dataset_opt['name'], len(test_set)))
test_loaders.append(test_loader)
model = ExtensibleTrainer(opt)
fea_loss = 0
2020-10-27 16:25:42 +00:00
psnr_loss = 0
for test_loader in test_loaders:
test_set_name = test_loader.dataset.opt['name']
logger.info('\nTesting [{:s}]...'.format(test_set_name))
test_start_time = time.time()
dataset_dir = osp.join(opt['path']['results_root'], test_set_name)
util.mkdir(dataset_dir)
test_results = OrderedDict()
test_results['psnr'] = []
test_results['ssim'] = []
test_results['psnr_y'] = []
test_results['ssim_y'] = []
tq = tqdm(test_loader)
for data in tq:
need_GT = False if test_loader.dataset.opt['dataroot_GT'] is None else True
2020-11-29 19:21:31 +00:00
need_GT = need_GT and want_metrics
2021-06-11 21:31:10 +00:00
fea_loss, psnr_loss = forward_pass(model, data, dataset_dir, opt)
2020-11-27 03:30:55 +00:00
fea_loss += fea_loss
psnr_loss += psnr_loss
# log
2020-10-27 16:25:42 +00:00
logger.info('# Validation # Fea: {:.4e}, PSNR: {:.4e}'.format(fea_loss / len(test_loader), psnr_loss / len(test_loader)))