[all] Passed black on all python files

bob/ip/binseg/configs/datasets/rimoner3odtest.py

@@ -7,13 +7,10 @@ from bob.ip.binseg.data.binsegdataset import BinSegDataset
 
 #### Config ####
 
-transforms = Compose([  
-                        Pad((8,8,8,8))
-                        ,ToTensor()
-                    ])
+transforms = Compose([Pad((8, 8, 8, 8)), ToTensor()])
 
 # bob.db.dataset init
-bobdb = RIMONER3(protocol = 'default_od')
+bobdb = RIMONER3(protocol="default_od")
 
 # PyTorch dataset
-dataset = BinSegDataset(bobdb, split='test', transform=transforms)
\ No newline at end of file
+dataset = BinSegDataset(bobdb, split="test", transform=transforms)

bob/ip/binseg/configs/datasets/stare.py

@@ -7,17 +7,19 @@ from bob.ip.binseg.data.binsegdataset import BinSegDataset
 
 #### Config ####
 
-transforms = Compose([  
-                        Pad((2,1,2,2))
-                        ,RandomHFlip()
-                        ,RandomVFlip()
-                        ,RandomRotation()
-                        ,ColorJitter()
-                        ,ToTensor()
-                    ])
+transforms = Compose(
+    [
+        Pad((2, 1, 2, 2)),
+        RandomHFlip(),
+        RandomVFlip(),
+        RandomRotation(),
+        ColorJitter(),
+        ToTensor(),
+    ]
+)
 
 # bob.db.dataset init
-bobdb = STARE(protocol = 'default')
+bobdb = STARE(protocol="default")
 
 # PyTorch dataset
-dataset = BinSegDataset(bobdb, split='train', transform=transforms)
\ No newline at end of file
+dataset = BinSegDataset(bobdb, split="train", transform=transforms)

bob/ip/binseg/configs/datasets/stare1024.py

@@ -7,19 +7,21 @@ from bob.ip.binseg.data.binsegdataset import BinSegDataset
 
 #### Config ####
 
-transforms = Compose([  
-                        RandomRotation()
-                        ,Pad((0,32,0,32))
-                        ,Resize(1024)
-                        ,CenterCrop(1024)
-                        ,RandomHFlip()
-                        ,RandomVFlip()
-                        ,ColorJitter()
-                        ,ToTensor()
-                    ])
+transforms = Compose(
+    [
+        RandomRotation(),
+        Pad((0, 32, 0, 32)),
+        Resize(1024),
+        CenterCrop(1024),
+        RandomHFlip(),
+        RandomVFlip(),
+        ColorJitter(),
+        ToTensor(),
+    ]
+)
 
 # bob.db.dataset init
-bobdb = STARE(protocol = 'default')
+bobdb = STARE(protocol="default")
 
 # PyTorch dataset
-dataset = BinSegDataset(bobdb, split='train', transform=transforms)
\ No newline at end of file
+dataset = BinSegDataset(bobdb, split="train", transform=transforms)

bob/ip/binseg/configs/datasets/stare1168.py

@@ -7,19 +7,21 @@ from bob.ip.binseg.data.binsegdataset import BinSegDataset
 
 #### Config ####
 
-transforms = Compose([  
-                        RandomRotation()
-                        ,Crop(50,0,500,705)
-                        ,Resize(1168)
-                        ,Pad((1,0,1,0))
-                        ,RandomHFlip()
-                        ,RandomVFlip()
-                        ,ColorJitter()
-                        ,ToTensor()
-                    ])
+transforms = Compose(
+    [
+        RandomRotation(),
+        Crop(50, 0, 500, 705),
+        Resize(1168),
+        Pad((1, 0, 1, 0)),
+        RandomHFlip(),
+        RandomVFlip(),
+        ColorJitter(),
+        ToTensor(),
+    ]
+)
 
 # bob.db.dataset init
-bobdb = STARE(protocol = 'default')
+bobdb = STARE(protocol="default")
 
 # PyTorch dataset
-dataset = BinSegDataset(bobdb, split='train', transform=transforms)
\ No newline at end of file
+dataset = BinSegDataset(bobdb, split="train", transform=transforms)

bob/ip/binseg/configs/datasets/stare544.py

@@ -7,17 +7,20 @@ from bob.ip.binseg.data.binsegdataset import BinSegDataset
 
 #### Config ####
 
-transforms = Compose([  RandomRotation()
-                        ,Resize(471)
-                        ,Pad((0,37,0,36))
-                        ,RandomHFlip()
-                        ,RandomVFlip()
-                        ,ColorJitter()
-                        ,ToTensor()
-                    ])
+transforms = Compose(
+    [
+        RandomRotation(),
+        Resize(471),
+        Pad((0, 37, 0, 36)),
+        RandomHFlip(),
+        RandomVFlip(),
+        ColorJitter(),
+        ToTensor(),
+    ]
+)
 
 # bob.db.dataset init
-bobdb = STARE(protocol = 'default')
+bobdb = STARE(protocol="default")
 
 # PyTorch dataset
-dataset = BinSegDataset(bobdb, split='train', transform=transforms)
\ No newline at end of file
+dataset = BinSegDataset(bobdb, split="train", transform=transforms)

bob/ip/binseg/configs/datasets/stare960.py

@@ -7,19 +7,21 @@ from bob.ip.binseg.data.binsegdataset import BinSegDataset
 
 #### Config ####
 
-transforms = Compose([  
-                        RandomRotation()
-                        ,Pad((0,32,0,32))
-                        ,Resize(960)
-                        ,CenterCrop(960)
-                        ,RandomHFlip()
-                        ,RandomVFlip()
-                        ,ColorJitter()
-                        ,ToTensor()
-                    ])
+transforms = Compose(
+    [
+        RandomRotation(),
+        Pad((0, 32, 0, 32)),
+        Resize(960),
+        CenterCrop(960),
+        RandomHFlip(),
+        RandomVFlip(),
+        ColorJitter(),
+        ToTensor(),
+    ]
+)
 
 # bob.db.dataset init
-bobdb = STARE(protocol = 'default')
+bobdb = STARE(protocol="default")
 
 # PyTorch dataset
-dataset = BinSegDataset(bobdb, split='train', transform=transforms)
\ No newline at end of file
+dataset = BinSegDataset(bobdb, split="train", transform=transforms)

bob/ip/binseg/configs/datasets/starechasedb1iostarhrf544.py

@@ -7,4 +7,4 @@ import torch
 #### Config ####
 
 # PyTorch dataset
-dataset = torch.utils.data.ConcatDataset([stare,chase,hrf,iostar])
\ No newline at end of file
+dataset = torch.utils.data.ConcatDataset([stare, chase, hrf, iostar])

bob/ip/binseg/configs/datasets/starechasedb1iostarhrf544ssldrive.py

@@ -5,30 +5,38 @@ from bob.ip.binseg.configs.datasets.hrf544 import dataset as hrf
 from bob.db.drive import Database as DRIVE
 from bob.ip.binseg.data.transforms import *
 import torch
-from bob.ip.binseg.data.binsegdataset import BinSegDataset, SSLBinSegDataset, UnLabeledBinSegDataset
+from bob.ip.binseg.data.binsegdataset import (
+    BinSegDataset,
+    SSLBinSegDataset,
+    UnLabeledBinSegDataset,
+)
 
 
 #### Config ####
 
 # PyTorch dataset
-labeled_dataset = torch.utils.data.ConcatDataset([stare,chase,iostar,hrf])
+labeled_dataset = torch.utils.data.ConcatDataset([stare, chase, iostar, hrf])
 
 #### Unlabeled STARE TRAIN ####
-unlabeled_transforms = Compose([  
-                        CenterCrop((544,544))
-                        ,RandomHFlip()
-                        ,RandomVFlip()
-                        ,RandomRotation()
-                        ,ColorJitter()
-                        ,ToTensor()
-                    ])
+unlabeled_transforms = Compose(
+    [
+        CenterCrop((544, 544)),
+        RandomHFlip(),
+        RandomVFlip(),
+        RandomRotation(),
+        ColorJitter(),
+        ToTensor(),
+    ]
+)
 
 # bob.db.dataset init
-drivebobdb = DRIVE(protocol = 'default')
+drivebobdb = DRIVE(protocol="default")
 
 # PyTorch dataset
-unlabeled_dataset = UnLabeledBinSegDataset(drivebobdb, split='train', transform=unlabeled_transforms)
+unlabeled_dataset = UnLabeledBinSegDataset(
+    drivebobdb, split="train", transform=unlabeled_transforms
+)
 
 # SSL Dataset
 
-dataset = SSLBinSegDataset(labeled_dataset, unlabeled_dataset)
\ No newline at end of file
+dataset = SSLBinSegDataset(labeled_dataset, unlabeled_dataset)

bob/ip/binseg/configs/datasets/staretest.py

@@ -7,13 +7,10 @@ from bob.ip.binseg.data.binsegdataset import BinSegDataset
 
 #### Config ####
 
-transforms = Compose([  
-                        Pad((2,1,2,2))
-                        ,ToTensor()
-                    ])
+transforms = Compose([Pad((2, 1, 2, 2)), ToTensor()])
 
 # bob.db.dataset init
-bobdb = STARE(protocol = 'default')
+bobdb = STARE(protocol="default")
 
 # PyTorch dataset
-dataset = BinSegDataset(bobdb, split='test', transform=transforms)
\ No newline at end of file
+dataset = BinSegDataset(bobdb, split="test", transform=transforms)

bob/ip/binseg/configs/models/driubn.py

@@ -26,13 +26,23 @@ scheduler_gamma = 0.1
 model = build_driu()
 
 # pretrained backbone
-pretrained_backbone = modelurls['vgg16_bn']
+pretrained_backbone = modelurls["vgg16_bn"]
 
 # optimizer
-optimizer = AdaBound(model.parameters(), lr=lr, betas=betas, final_lr=final_lr, gamma=gamma,
-                 eps=eps, weight_decay=weight_decay, amsbound=amsbound) 
+optimizer = AdaBound(
+    model.parameters(),
+    lr=lr,
+    betas=betas,
+    final_lr=final_lr,
+    gamma=gamma,
+    eps=eps,
+    weight_decay=weight_decay,
+    amsbound=amsbound,
+)
 # criterion
 criterion = SoftJaccardBCELogitsLoss(alpha=0.7)
 
 # scheduler
-scheduler = MultiStepLR(optimizer, milestones=scheduler_milestones, gamma=scheduler_gamma)
+scheduler = MultiStepLR(
+    optimizer, milestones=scheduler_milestones, gamma=scheduler_gamma
+)

bob/ip/binseg/configs/models/driubnssl.py

@@ -26,14 +26,24 @@ scheduler_gamma = 0.1
 model = build_driu()
 
 # pretrained backbone
-pretrained_backbone = modelurls['vgg16_bn']
+pretrained_backbone = modelurls["vgg16_bn"]
 
 # optimizer
-optimizer = AdaBound(model.parameters(), lr=lr, betas=betas, final_lr=final_lr, gamma=gamma,
-                 eps=eps, weight_decay=weight_decay, amsbound=amsbound) 
+optimizer = AdaBound(
+    model.parameters(),
+    lr=lr,
+    betas=betas,
+    final_lr=final_lr,
+    gamma=gamma,
+    eps=eps,
+    weight_decay=weight_decay,
+    amsbound=amsbound,
+)
 
 # criterion
 criterion = MixJacLoss(lambda_u=0.05, jacalpha=0.7)
 
 # scheduler
-scheduler = MultiStepLR(optimizer, milestones=scheduler_milestones, gamma=scheduler_gamma)
+scheduler = MultiStepLR(
+    optimizer, milestones=scheduler_milestones, gamma=scheduler_gamma
+)

bob/ip/binseg/configs/models/driuod.py

@@ -26,13 +26,23 @@ scheduler_gamma = 0.1
 model = build_driuod()
 
 # pretrained backbone
-pretrained_backbone = modelurls['vgg16']
+pretrained_backbone = modelurls["vgg16"]
 
 # optimizer
-optimizer = AdaBound(model.parameters(), lr=lr, betas=betas, final_lr=final_lr, gamma=gamma,
-                 eps=eps, weight_decay=weight_decay, amsbound=amsbound) 
+optimizer = AdaBound(
+    model.parameters(),
+    lr=lr,
+    betas=betas,
+    final_lr=final_lr,
+    gamma=gamma,
+    eps=eps,
+    weight_decay=weight_decay,
+    amsbound=amsbound,
+)
 # criterion
 criterion = SoftJaccardBCELogitsLoss(alpha=0.7)
 
 # scheduler
-scheduler = MultiStepLR(optimizer, milestones=scheduler_milestones, gamma=scheduler_gamma)
+scheduler = MultiStepLR(
+    optimizer, milestones=scheduler_milestones, gamma=scheduler_gamma
+)

bob/ip/binseg/configs/models/driussl.py

@@ -26,14 +26,24 @@ scheduler_gamma = 0.1
 model = build_driu()
 
 # pretrained backbone
-pretrained_backbone = modelurls['vgg16']
+pretrained_backbone = modelurls["vgg16"]
 
 # optimizer
-optimizer = AdaBound(model.parameters(), lr=lr, betas=betas, final_lr=final_lr, gamma=gamma,
-                 eps=eps, weight_decay=weight_decay, amsbound=amsbound) 
+optimizer = AdaBound(
+    model.parameters(),
+    lr=lr,
+    betas=betas,
+    final_lr=final_lr,
+    gamma=gamma,
+    eps=eps,
+    weight_decay=weight_decay,
+    amsbound=amsbound,
+)
 
 # criterion
 criterion = MixJacLoss(lambda_u=0.05, jacalpha=0.7)
 
 # scheduler
-scheduler = MultiStepLR(optimizer, milestones=scheduler_milestones, gamma=scheduler_gamma)
+scheduler = MultiStepLR(
+    optimizer, milestones=scheduler_milestones, gamma=scheduler_gamma
+)

bob/ip/binseg/configs/models/hed.py

@@ -27,13 +27,23 @@ scheduler_gamma = 0.1
 model = build_hed()
 
 # pretrained backbone
-pretrained_backbone = modelurls['vgg16']
+pretrained_backbone = modelurls["vgg16"]
 
 # optimizer
-optimizer = AdaBound(model.parameters(), lr=lr, betas=betas, final_lr=final_lr, gamma=gamma,
-                 eps=eps, weight_decay=weight_decay, amsbound=amsbound) 
+optimizer = AdaBound(
+    model.parameters(),
+    lr=lr,
+    betas=betas,
+    final_lr=final_lr,
+    gamma=gamma,
+    eps=eps,
+    weight_decay=weight_decay,
+    amsbound=amsbound,
+)
 # criterion
 criterion = HEDSoftJaccardBCELogitsLoss(alpha=0.7)
 
 # scheduler
-scheduler = MultiStepLR(optimizer, milestones=scheduler_milestones, gamma=scheduler_gamma)
+scheduler = MultiStepLR(
+    optimizer, milestones=scheduler_milestones, gamma=scheduler_gamma
+)

bob/ip/binseg/configs/models/m2unet.py

@@ -26,14 +26,24 @@ scheduler_gamma = 0.1
 model = build_m2unet()
 
 # pretrained backbone
-pretrained_backbone = modelurls['mobilenetv2']
+pretrained_backbone = modelurls["mobilenetv2"]
 
 # optimizer
-optimizer = AdaBound(model.parameters(), lr=lr, betas=betas, final_lr=final_lr, gamma=gamma,
-                 eps=eps, weight_decay=weight_decay, amsbound=amsbound) 
-    
+optimizer = AdaBound(
+    model.parameters(),
+    lr=lr,
+    betas=betas,
+    final_lr=final_lr,
+    gamma=gamma,
+    eps=eps,
+    weight_decay=weight_decay,
+    amsbound=amsbound,
+)
+
 # criterion
 criterion = SoftJaccardBCELogitsLoss(alpha=0.7)
 
 # scheduler
-scheduler = MultiStepLR(optimizer, milestones=scheduler_milestones, gamma=scheduler_gamma)
+scheduler = MultiStepLR(
+    optimizer, milestones=scheduler_milestones, gamma=scheduler_gamma
+)

bob/ip/binseg/configs/models/m2unetssl.py

@@ -26,14 +26,24 @@ scheduler_gamma = 0.1
 model = build_m2unet()
 
 # pretrained backbone
-pretrained_backbone = modelurls['mobilenetv2']
+pretrained_backbone = modelurls["mobilenetv2"]
 
 # optimizer
-optimizer = AdaBound(model.parameters(), lr=lr, betas=betas, final_lr=final_lr, gamma=gamma,
-                 eps=eps, weight_decay=weight_decay, amsbound=amsbound) 
-    
+optimizer = AdaBound(
+    model.parameters(),
+    lr=lr,
+    betas=betas,
+    final_lr=final_lr,
+    gamma=gamma,
+    eps=eps,
+    weight_decay=weight_decay,
+    amsbound=amsbound,
+)
+
 # criterion
 criterion = MixJacLoss(lambda_u=0.05, jacalpha=0.7)
 
 # scheduler
-scheduler = MultiStepLR(optimizer, milestones=scheduler_milestones, gamma=scheduler_gamma)
+scheduler = MultiStepLR(
+    optimizer, milestones=scheduler_milestones, gamma=scheduler_gamma
+)

bob/ip/binseg/configs/models/resunet.py

@@ -26,14 +26,24 @@ scheduler_gamma = 0.1
 model = build_res50unet()
 
 # pretrained backbone
-pretrained_backbone = modelurls['resnet50']
+pretrained_backbone = modelurls["resnet50"]
 
 # optimizer
-optimizer = AdaBound(model.parameters(), lr=lr, betas=betas, final_lr=final_lr, gamma=gamma,
-                 eps=eps, weight_decay=weight_decay, amsbound=amsbound) 
-    
+optimizer = AdaBound(
+    model.parameters(),
+    lr=lr,
+    betas=betas,
+    final_lr=final_lr,
+    gamma=gamma,
+    eps=eps,
+    weight_decay=weight_decay,
+    amsbound=amsbound,
+)
+
 # criterion
 criterion = SoftJaccardBCELogitsLoss(alpha=0.7)
 
 # scheduler
-scheduler = MultiStepLR(optimizer, milestones=scheduler_milestones, gamma=scheduler_gamma)
+scheduler = MultiStepLR(
+    optimizer, milestones=scheduler_milestones, gamma=scheduler_gamma
+)

bob/ip/binseg/configs/models/unet.py

@@ -26,14 +26,24 @@ scheduler_gamma = 0.1
 model = build_unet()
 
 # pretrained backbone
-pretrained_backbone = modelurls['vgg16']
+pretrained_backbone = modelurls["vgg16"]
 
 # optimizer
-optimizer = AdaBound(model.parameters(), lr=lr, betas=betas, final_lr=final_lr, gamma=gamma,
-                 eps=eps, weight_decay=weight_decay, amsbound=amsbound) 
-    
+optimizer = AdaBound(
+    model.parameters(),
+    lr=lr,
+    betas=betas,
+    final_lr=final_lr,
+    gamma=gamma,
+    eps=eps,
+    weight_decay=weight_decay,
+    amsbound=amsbound,
+)
+
 # criterion
 criterion = SoftJaccardBCELogitsLoss(alpha=0.7)
 
 # scheduler
-scheduler = MultiStepLR(optimizer, milestones=scheduler_milestones, gamma=scheduler_gamma)
+scheduler = MultiStepLR(
+    optimizer, milestones=scheduler_milestones, gamma=scheduler_gamma
+)

bob/ip/binseg/data/binsegdataset.py

@@ -3,6 +3,7 @@
 from torch.utils.data import Dataset
 import random
 
+
 class BinSegDataset(Dataset):
     """PyTorch dataset wrapper around bob.db binary segmentation datasets. 
     A transform object can be passed that will be applied to the image, ground truth and mask (if present). 
@@ -19,18 +20,19 @@ class BinSegDataset(Dataset):
     mask : bool
         whether dataset contains masks or not
     """
-    def __init__(self, bobdb, split = 'train', transform = None,index_to = None):
+
+    def __init__(self, bobdb, split="train", transform=None, index_to=None):
         if index_to:
             self.database = bobdb.samples(split)[:index_to]
         else:
             self.database = bobdb.samples(split)
         self.transform = transform
         self.split = split
-    
+
     @property
     def mask(self):
         # check if first sample contains a mask
-        return hasattr(self.database[0], 'mask')
+        return hasattr(self.database[0], "mask")
 
     def __len__(self):
         """
@@ -40,8 +42,8 @@ class BinSegDataset(Dataset):
             size of the dataset
         """
         return len(self.database)
-    
-    def __getitem__(self,index):
+
+    def __getitem__(self, index):
         """
         Parameters
         ----------
@@ -56,12 +58,12 @@ class BinSegDataset(Dataset):
         gt = self.database[index].gt.pil_image()
         img_name = self.database[index].img.basename
         sample = [img, gt]
-        
-        if self.transform :
+
+        if self.transform:
             sample = self.transform(*sample)
-        
-        sample.insert(0,img_name)
-        
+
+        sample.insert(0, img_name)
+
         return sample
 
 
@@ -77,10 +79,10 @@ class SSLBinSegDataset(Dataset):
     unlabeled_dataset : :py:class:`torch.utils.data.Dataset`
         UnLabeledBinSegDataset with unlabeled data
     """
+
     def __init__(self, labeled_dataset, unlabeled_dataset):
         self.labeled_dataset = labeled_dataset
         self.unlabeled_dataset = unlabeled_dataset
-    
 
     def __len__(self):
         """
@@ -90,8 +92,8 @@ class SSLBinSegDataset(Dataset):
             size of the dataset
         """
         return len(self.labeled_dataset)
-    
-    def __getitem__(self,index):
+
+    def __getitem__(self, index):
         """
         Parameters
         ----------
@@ -123,13 +125,14 @@ class UnLabeledBinSegDataset(Dataset):
     transform : :py:mod:`bob.ip.binseg.data.transforms`, optional
         A transform or composition of transfroms. Defaults to ``None``.
     """
-    def __init__(self, db, split = 'train', transform = None,index_from= None):
+
+    def __init__(self, db, split="train", transform=None, index_from=None):
         if index_from:
             self.database = db.samples(split)[index_from:]
         else:
             self.database = db.samples(split)
         self.transform = transform
-        self.split = split   
+        self.split = split
 
     def __len__(self):
         """
@@ -139,8 +142,8 @@ class UnLabeledBinSegDataset(Dataset):
             size of the dataset
         """
         return len(self.database)
-    
-    def __getitem__(self,index):
+
+    def __getitem__(self, index):
         """
         Parameters
         ----------
@@ -155,9 +158,9 @@ class UnLabeledBinSegDataset(Dataset):
         img = self.database[index].img.pil_image()
         img_name = self.database[index].img.basename
         sample = [img]
-        if self.transform :
+        if self.transform:
             sample = self.transform(img)
-        
-        sample.insert(0,img_name)
-        
-        return sample
\ No newline at end of file
+
+        sample.insert(0, img_name)
+
+        return sample

bob/ip/binseg/data/imagefolder.py

@@ -8,16 +8,18 @@ import torch
 import torchvision.transforms.functional as VF
 import bob.io.base
 
+
 def get_file_lists(data_path):
     data_path = Path(data_path)
-    
-    image_path = data_path.joinpath('images')
-    image_file_names = np.array(sorted(list(image_path.glob('*'))))
 
-    gt_path = data_path.joinpath('gt')
-    gt_file_names = np.array(sorted(list(gt_path.glob('*'))))
+    image_path = data_path.joinpath("images")
+    image_file_names = np.array(sorted(list(image_path.glob("*"))))
+
+    gt_path = data_path.joinpath("gt")
+    gt_file_names = np.array(sorted(list(gt_path.glob("*"))))
     return image_file_names, gt_file_names
 
+
 class ImageFolder(Dataset):
     """
     Generic ImageFolder dataset, that contains two folders:
@@ -32,7 +34,8 @@ class ImageFolder(Dataset):
         full path to root of dataset
     
     """
-    def __init__(self, path, transform = None):
+
+    def __init__(self, path, transform=None):
         self.transform = transform
         self.img_file_list, self.gt_file_list = get_file_lists(path)
 
@@ -44,8 +47,8 @@ class ImageFolder(Dataset):
             size of the dataset
         """
         return len(self.img_file_list)
-    
-    def __getitem__(self,index):
+
+    def __getitem__(self, index):
         """
         Parameters
         ----------
@@ -58,22 +61,22 @@ class ImageFolder(Dataset):
         """
         img_path = self.img_file_list[index]
         img_name = img_path.name
-        img = Image.open(img_path).convert(mode='RGB')
-    
+        img = Image.open(img_path).convert(mode="RGB")
+
         gt_path = self.gt_file_list[index]
-        if gt_path.suffix == '.hdf5':
-            gt = bob.io.base.load(str(gt_path)).astype('float32')
+        if gt_path.suffix == ".hdf5":
+            gt = bob.io.base.load(str(gt_path)).astype("float32")
             # not elegant but since transforms require PIL images we do this hacky workaround here
             gt = torch.from_numpy(gt)
-            gt = VF.to_pil_image(gt).convert(mode='1', dither=None)
+            gt = VF.to_pil_image(gt).convert(mode="1", dither=None)
         else:
-            gt = Image.open(gt_path).convert(mode='1', dither=None)
-        
+            gt = Image.open(gt_path).convert(mode="1", dither=None)
+
         sample = [img, gt]
-        
-        if self.transform :
+
+        if self.transform:
             sample = self.transform(*sample)
-        
-        sample.insert(0,img_name)
-        
+
+        sample.insert(0, img_name)
+
         return sample