torch.set_grad_enabled(False) # make sure to not compute gradients for computational performance
torch.backends.cudnn.enabled = True
device = torch.device("cuda" if torch.cuda.is_available() else "cpu") # either gpu or cpu
dataset_name = 'CPC'
pairs_which_dataset = pd.read_csv(dataset_name + '/pairs_which_dataset.txt')
pairs_gts = np.loadtxt(dataset_name + '/pairs_with_gt.txt')
pairs_gts = pd.DataFrame(pairs_gts)
pairs = pairs_gts[pairs_gts.columns[0:2]]
l_pairs = pairs[pairs.columns[0]]
r_pairs = pairs[pairs.columns[1]]
size = np.size(pairs_gts[0])
matches = {'Matches': []}
with torch.no_grad():
network = GLU_Net(path_pre_trained_models=args.pre_trained_models_dir,
model_type=args.pre_trained_model,
consensus_network=False,
cyclic_consistency=True,
iterative_refinement=True,
apply_flipping_condition=False)
for idx in range(size):
print(idx)
l = "{:08}".format(int(l_pairs[idx]))
r = "{:08}".format(int(r_pairs[idx]))
folder = pairs_which_dataset.iloc[idx]
I1 = dataset_name + '/' + folder[0] + 'Images/' + l + '.jpg'
I2 = dataset_name + '/' + folder[0] + 'Images/' + r + '.jpg'
# I2 = imageio.imread(pairs_which_dataset(idx) + 'Images/' sprintf('%.8d.jpg', r)]);
try:
source_image = imageio.imread(I1)
target_image = imageio.imread(I2)
source_image, target_image = pad_to_same_shape(source_image, target_image)
# define the image processing parameters, the actual pre-processing is done within the model functions
input_images_transform = transforms.Compose([ArrayToTensor(get_float=False)
]) # only put channel first
gt_flow_transform = transforms.Compose([ArrayToTensor()
]) # only put channel first
co_transform = None
for pre_trained_model_type in pre_trained_models:
print(pre_trained_model_type)
with torch.no_grad():
# define the network to use
if args.model == 'GLUNet':
network = GLU_Net(model_type=pre_trained_model_type,
consensus_network=False,
cyclic_consistency=True,
iterative_refinement=True,
apply_flipping_condition=args.flipping_condition)
elif args.model == 'SemanticGLUNet':
network = GLU_Net(model_type=pre_trained_model_type,
feature_concatenation=True,
cyclic_consistency=False,
consensus_network=True,
iterative_refinement=True,
apply_flipping_condition=args.flipping_condition)
elif args.model == 'GLOCALNet':
network = GLOCAL_Net(model_type=pre_trained_model_type,
constrained_corr=True,
global_corr=True)
# ETH3D dataset information
dataset_names = [
'lakeside', 'sand_box', 'storage_room', 'storage_room_2', 'tunnel',
'delivery_area', 'electro', 'forest', 'playground', 'terrains'
]
rates = list(range(3, 16, 2))
for pre_trained_model_type in pre_trained_models:
print('model: ' + args.model + ', pre-trained model: ' +
pre_trained_model_type)
with torch.no_grad():
# define the network to use
if args.model == 'GLUNet':
network = GLU_Net(model_type=pre_trained_model_type,
consensus_network=False,
cyclic_consistency=True,
iterative_refinement=True)
elif args.model == 'SemanticGLUNet':
network = GLU_Net(model_type=pre_trained_model_type,
feature_concatenation=True,
cyclic_consistency=False,
consensus_network=True,
iterative_refinement=True,
apply_flipping_condition=args.flipping_condition)
elif args.model == 'GLOCALNet':
network = GLOCAL_Net(model_type=pre_trained_model_type,
constrained_corr=True,
global_corr=True)