Fix minor bug in new data provider

data.py

@@ -19,8 +19,9 @@ def augment_image(image, k, channels):
 
 class MatchingNetworkDatasetParallel(Dataset):
     def __init__(self, batch_size, reverse_channels, num_of_gpus, image_height, image_width, image_channels,
-                 train_val_test_split, num_classes_per_set, num_samples_per_class, seed=100,
-                 reset_stored_filepaths=False, labels_as_int=False):
+                 train_val_test_split, num_classes_per_set, num_samples_per_class,
+                 data_path, dataset_name,  indexes_of_folders_indicating_class,
+                 seed=100, reset_stored_filepaths=False, labels_as_int=False):
         """
         :param batch_size: The batch size to use for the data loader
         :param last_training_class_index: The final index for the training set, used to restrict the training set
@@ -30,6 +31,9 @@ def __init__(self, batch_size, reverse_channels, num_of_gpus, image_height, imag
         :param num_of_gpus: Number of gpus to use for training
         :param gen_batches: How many batches to use from the validation set for the end of epoch generations
         """
+        self.data_path = data_path
+        self.dataset_name = dataset_name
+        self.indexes_of_folders_indicating_class = indexes_of_folders_indicating_class
         self.labels_as_int = labels_as_int
         self.train_val_test_split = train_val_test_split
         self.current_dataset_name = "train"
@@ -171,7 +175,11 @@ def get_label_from_index(self, index):
         return index_to_label_name[index]
 
     def get_label_from_path(self, filepath):
-        raise NotImplementedError
+        label_bits = filepath.split("/")
+        label = "_".join([label_bits[idx] for idx in self.indexes_of_folders_indicating_class])
+        if self.labels_as_int:
+            label = int(label)
+        return label
 
     def load_image(self, image_path, channels):
 
@@ -418,36 +426,27 @@ def sample_iter_data(self, sample, num_gpus, batch_size, samples_per_iter):
 
 class FolderMatchingNetworkDatasetParallel(MatchingNetworkDatasetParallel):
     def __init__(self, name, num_of_gpus, batch_size, image_height, image_width, image_channels,
-                 train_val_test_split, data_path, index_of_folder_indicating_class, reset_stored_filepaths,
+                 train_val_test_split, data_path, indexes_of_folders_indicating_class, reset_stored_filepaths,
                  num_samples_per_class, num_classes_per_set, labels_as_int, reverse_channels):
 
-        self.data_path = os.path.abspath(data_path)
-        self.dataset_name = name
-        self.indeces_of_folders_indicating_class_list = index_of_folder_indicating_class
-
         super(FolderMatchingNetworkDatasetParallel, self).__init__(
             batch_size=batch_size, reverse_channels=reverse_channels,
             num_of_gpus=num_of_gpus, image_height=image_height,
             image_width=image_width, image_channels=image_channels,
             train_val_test_split=train_val_test_split, reset_stored_filepaths=reset_stored_filepaths,
             num_classes_per_set=num_classes_per_set, num_samples_per_class=num_samples_per_class,
-            labels_as_int=labels_as_int)
+            labels_as_int=labels_as_int, data_path=os.path.abspath(data_path), dataset_name=name,
+            indexes_of_folders_indicating_class=indexes_of_folders_indicating_class)
 
-    def get_label_from_path(self, filepath):
-        label_bits = filepath.split("/")
-        label = "_".join([label_bits[idx] for idx in self.indeces_of_folders_indicating_class_list])
-        if self.labels_as_int:
-            label = int(label)
-        return label
 
 class FolderDatasetLoader(MatchingNetworkLoader):
     def __init__(self, name, batch_size, image_height, image_width, image_channels, data_path, train_val_test_split,
-                 num_of_gpus=1, samples_per_iter=1, num_workers=4, index_of_folder_indicating_class=[-2, -3],
+                 num_of_gpus=1, samples_per_iter=1, num_workers=4, indexes_of_folders_indicating_class=[-2],
                  reset_stored_filepaths=False, num_samples_per_class=1, num_classes_per_set=20, reverse_channels=False,
                  seed=100, label_as_int=False):
 
         self.name = name
-        self.index_of_folder_indicating_class = index_of_folder_indicating_class
+        self.indexes_of_folders_indicating_class = indexes_of_folders_indicating_class
         self.reset_stored_filepaths = reset_stored_filepaths
         super(FolderDatasetLoader, self).__init__(name, num_of_gpus, batch_size, image_height, image_width, image_channels, num_classes_per_set, data_path,
                  num_samples_per_class, train_val_test_split,
@@ -457,11 +456,11 @@ def get_dataset(self, batch_size, reverse_channels, num_of_gpus, image_height, i
                  train_val_test_split, num_classes_per_set, num_samples_per_class, seed,
                  reset_stored_filepaths, data_path, labels_as_int):
         return FolderMatchingNetworkDatasetParallel(name=self.name, num_of_gpus=num_of_gpus, batch_size=batch_size,
-                                              image_height=image_height, image_width=image_width,
-                                              image_channels=image_channels,
-                                              train_val_test_split=train_val_test_split, data_path=data_path,
-                                              index_of_folder_indicating_class=self.index_of_folder_indicating_class,
-                                              reset_stored_filepaths=self.reset_stored_filepaths,
-                                              num_samples_per_class=num_samples_per_class,
-                                              num_classes_per_set=num_classes_per_set, labels_as_int=labels_as_int,
-                                              reverse_channels=reverse_channels)
+                                                    image_height=image_height, image_width=image_width,
+                                                    image_channels=image_channels,
+                                                    train_val_test_split=train_val_test_split, data_path=data_path,
+                                                    indexes_of_folders_indicating_class=self.indexes_of_folders_indicating_class,
+                                                    reset_stored_filepaths=self.reset_stored_filepaths,
+                                                    num_samples_per_class=num_samples_per_class,
+                                                    num_classes_per_set=num_classes_per_set, labels_as_int=labels_as_int,
+                                                    reverse_channels=reverse_channels)

train_one_shot_learning_matching_network.py

@@ -15,7 +15,7 @@
                                    train_val_test_split=(1200/1622, 211/1622, 211/1622),
                                    samples_per_iter=1, num_workers=4,
                                    data_path="datasets/omniglot_dataset", name="omniglot_dataset",
-                                   index_of_folder_indicating_class=[-2, -3], reset_stored_filepaths=True,
+                                   indexes_of_folders_indicating_class=[-2, -3], reset_stored_filepaths=True,
                                    num_samples_per_class=args.samples_per_class,
                                    num_classes_per_set=args.classes_per_set, label_as_int=False)