|
import os,time |
|
os.environ["KMP_DUPLICATE_LIB_OK"]="TRUE" |
|
|
|
from options.test_options import TestOptions |
|
from dataloader.data_loader import dataloader |
|
from model import create_model |
|
from itertools import islice |
|
from util.visualizer import save_images |
|
from util import html |
|
|
|
if __name__=='__main__': |
|
opt = TestOptions().parse() |
|
opt.name = 'imagenet' |
|
opt.img_file='../../tmp/img/' |
|
opt.mask_file='../../tmp/mask/' |
|
opt.results_dir='../../results' |
|
opt.model='tc' |
|
opt.coarse_or_refine='refine' |
|
opt.gpu_id=0 |
|
opt.no_shuffle=True |
|
opt.batch_size=1 |
|
opt.preprocess='scale_shortside' |
|
opt.mask_type=3 |
|
opt.load_size=512 |
|
opt.attn_G=True |
|
opt.add_noise=True |
|
dataset = dataloader(opt) |
|
dataset_size = len(dataset) * opt.batch_size |
|
print('testing images = %d' % dataset_size) |
|
model = create_model(opt) |
|
|
|
opt.epoch = '%d' % opt.which_iter if opt.which_iter > 0 else opt.epoch |
|
web_dir = os.path.join(opt.results_dir, opt.name, '{}_{}'.format(opt.phase, opt.epoch)) |
|
print('creating web directory', web_dir) |
|
opt.save_dir = web_dir |
|
webpage = html.HTML(web_dir, 'Experiment = %s, Phase = %s, Epoch = %s' % (opt.name, opt.phase, opt.epoch)) |
|
opt.how_many = dataset_size if opt.how_many == float("inf") else opt.how_many |
|
|
|
iter_data_time = time.time() |
|
for i, data in enumerate(islice(dataset, opt.how_many)): |
|
if i == 0: |
|
model.setup(opt) |
|
model.parallelize() |
|
model.eval() |
|
model.set_input(data) |
|
model.test() |
|
visuals = model.get_current_visuals() |
|
img_path = model.get_image_paths() |
|
save_images(webpage, visuals, img_path, width=opt.display_winsize) |
|
if i % 5 == 0: |
|
print('processing (%04d)-th image... %s' % (i, img_path)) |
|
total_time = time.time() - iter_data_time |
|
print('the total evaluation time %f' % (total_time)) |
|
webpage.save() |
|
|