def __init__(self, root, split='train', mode=None):
super(Promise12, self).__init__(root, split=split, mode=mode)
self.mode = mode
#self.joint_transform = joint_transform
root = root + '/' + self.BASE_DIR
self.joint_transform = Compose([
RandomTranslate(offset=(0.2, 0.1)),
RandomVerticallyFlip(),
RandomHorizontallyFlip(),
RandomElasticTransform(alpha = 1.5, sigma = 0.07, img_type='F'),
])
self.img_normalize = None
# SECOND
# store data in the npy file
data_path = os.path.join(root, 'npy_image')
if not os.path.exists(data_path):
create_exp_dir(data_path, 'Create augmentation data at {}')
data_to_array(root, data_path, self.CROP_SIZE, self.CROP_SIZE)
else:
print('read the data from: {}'.format(data_path))
self.test_file_list = get_test_list(root)
# read the data from npy
if mode == 'train':
self.X_train, self.y_train = load_train_data(data_path)
self.size = self.X_train.shape[0]
elif mode == 'val':
self.X_val, self.y_val = load_val_data(data_path)
self.size = self.X_val.shape[0]
elif mode == 'test':
self.X_test, self.x_slice_array = load_test_data(data_path)
self.size = self.X_test.shape[0]
def main(format='svg'):
genotype_name = 'NAS_UNET_NEW_V2'
if len(sys.argv) != 2:
print('usage:\n python {} ARCH_NAME, Default: NAS_UNET_V2'.format(
sys.argv[0]))
else:
genotype_name = sys.argv[1]
store_path = './cell_visualize/' + '/{}'.format(format) + '/{}'.format(
genotype_name)
create_exp_dir(store_path)
if 'Windows' in platform.platform():
os.environ['PATH'] += os.pathsep + '../3rd_tools/graphviz-2.38/bin/'
try:
genotype = eval('geno_types.{}'.format(genotype_name))
except AttributeError:
print('{} is not specified in geno_types.py'.format(genotype_name))
sys.exit(1)
visualize.plot(genotype.down, store_path + '/DownC', format=format)
visualize.plot(genotype.up, store_path + '/UpC', format=format)
res = [[s + 1, l + 1] for s, l in zip(list(start), list(length))]
res = list(chain.from_iterable(res))
return ' '.join([str(r) for r in res])
if __name__ == '__main__':
input_path = '../../../predictions/nerve_rst'
masks = [f for f in os.listdir(input_path) if f.endswith('_mask.tif')]
masks = sorted(masks, key=lambda d: int(d.split('_')[0]))
encodings = []
total = len(masks)
for m in masks:
img = Image.open(os.path.join(input_path, m))
x = np.array(img.getdata(), dtype=np.uint8).reshape(img.size[::-1])
x = x // 255
print('----->: processing: {}'.format(m))
encodings.append(rle_encoding(x))
print('Encode done, write to submission file')
# check output
first_row = 'img,pixels'
create_exp_dir('./submission', '=> create submission file')
file_name = os.path.join('submission', 'submission.csv')
with open(file_name, 'w+') as f:
f.write(first_row + '\n')
for i in range(total):
s = str(i + 1) + ',' + encodings[i]
f.write(s + '\n')
f.close()
print('write done!')
def test(self, img_queue, split='val', desc=''):
self.model.eval()
predict_list = []
accuracy = 0
tbar = tqdm(img_queue)
create_exp_dir(desc, desc='=>Save prediction image on')
with torch.no_grad():
for step, (input, target) in enumerate(tbar):
input = input.cuda(self.device)
if not isinstance(target, list):
target = target.cuda(self.device)
predicts = self.model(input)
# for cityscapes, voc, camvid, test have label
if not isinstance(target, list):
test_loss = self.criterion(predicts[0], target)
self.loss_meter.update(test_loss.item())
self.metric.update(target, predicts[0])
if step % self.cfg['training']['report_freq'] == 0:
pixAcc, mIoU = self.metric.get()
self.logger.info(
'{} loss: {}, pixAcc: {}, mIoU: {}'.format(
split, self.loss_meter.mloss, pixAcc, mIoU))
tbar.set_description(
'loss: %.6f, pixAcc: %.3f, mIoU: %.6f' %
(self.loss_meter.mloss, pixAcc, mIoU))
accuracy += dice_coefficient(predicts[0].cpu(),
target.cpu())
else:
N = predicts[0].shape[0]
for i in range(N):
if self.args.crf: # use crf
predict = torch.argmax(predicts[0].cpu(), 1)[i]
predict = dense_crf(
np.array(input[i].cpu()).astype(np.uint8),
predict) > 0.5
img = Image.fromarray(
(predict * 255).astype(np.uint8))
file_name = os.path.split(target[i])[1]
file_name = file_name.split('.')[0] + '_mask.tif'
img.save(os.path.join(desc, file_name))
else:
img = Image.fromarray(
(torch.argmax(predicts[0].cpu(), 1)[i] *
255).numpy().astype(np.uint8))
file_name = os.path.split(target[i])[1]
file_name = file_name.split('.')[0] + '_mask.tif'
img.save(os.path.join(desc, file_name))
if desc == 'promise12': # for promise12, test have not label or have label to calc extra metric
predict_list += [
torch.argmax(predicts[0], dim=1).cpu().numpy()
]
print('==> accuracy: {}'.format(accuracy / len(img_queue)))
# cause the predicts is a list [pred, aux_pred(may not)]
if len(predicts[0].shape) == 4: #
pred = predicts[0]
else:
pred = predicts
# save images
if not isinstance(target, list) and not isinstance(target, str): #
grid_image = store_images(input, pred, target)
pixAcc, mIoU = self.metric.get()
self.logger.info('{}/loss: {}, pixAcc: {}, mIoU: {}'.format(
split, self.loss_meter.mloss, pixAcc, mIoU))
elif desc == 'promise12': # for promise12, test have not label
predict_test(predict_list, target,
self.save_path + '/{}_rst'.format(split))
# for promise12 metirc
if desc == 'promise12' and split == 'val':
val_list = [5, 15, 25, 35, 45]
dir = os.path.join(self.trainset.root_dir, self.trainset.base_dir,
'TrainingData')
biomedical_image_metric(predict_list, val_list, dir + '/')