Improve previous commit to avoid mutable default arguments

mrcnn/model.py

@@ -1637,7 +1637,7 @@ def generate_random_rois(image_shape, count, gt_class_ids, gt_boxes):
 
 def data_generator(dataset, config, shuffle=True, augment=False, augmentation=None,
                    random_rois=0, batch_size=1, detection_targets=False,
-                   no_augmentation_sources=[]):
+                   no_augmentation_sources=None):
     """A generator that returns images and corresponding target class ids,
     bounding box deltas, and masks.
 
@@ -1656,6 +1656,9 @@ def data_generator(dataset, config, shuffle=True, augment=False, augmentation=No
     detection_targets: If True, generate detection targets (class IDs, bbox
         deltas, and masks). Typically for debugging or visualizations because
         in trainig detection targets are generated by DetectionTargetLayer.
+    no_augmentation_sources: Optional. List of sources to exclude for
+        augmentation. A source is string that identifies a dataset and is
+        defined in the Dataset class.
 
     Returns a Python generator. Upon calling next() on it, the
     generator returns two lists, inputs and outputs. The containtes
@@ -1674,13 +1677,12 @@ def data_generator(dataset, config, shuffle=True, augment=False, augmentation=No
     outputs list: Usually empty in regular training. But if detection_targets
         is True then the outputs list contains target class_ids, bbox deltas,
         and masks.
-
-    no_augmentation_sources: (list) Optional. List of sources to be skipped for augmentation
     """
     b = 0  # batch item index
     image_index = -1
     image_ids = np.copy(dataset.image_ids)
     error_count = 0
+    no_augmentation_sources = no_augmentation_sources or []
 
     # Anchors
     # [anchor_count, (y1, x1, y2, x2)]
@@ -2283,7 +2285,7 @@ class MaskRCNN():
             "*epoch*", "{epoch:04d}")
 
     def train(self, train_dataset, val_dataset, learning_rate, epochs, layers,
-              augmentation=None, custom_callbacks=[], no_augmentation_sources=[]):
+              augmentation=None, custom_callbacks=None, no_augmentation_sources=None):
         """Train the model.
         train_dataset, val_dataset: Training and validation Dataset objects.
         learning_rate: The learning rate to train with
@@ -2310,12 +2312,11 @@ class MaskRCNN():
                     imgaug.augmenters.Fliplr(0.5),
                     imgaug.augmenters.GaussianBlur(sigma=(0.0, 5.0))
                 ])
-	    custom_callbacks: (list) Optional. Add custom callbacks to be called
+	    custom_callbacks: Optional. Add custom callbacks to be called
 	        with the keras fit_generator method. Must be list of type keras.callbacks. 
-
-        no_augmentation_sources: (list) Optional. List of sources to be skipped for augmentation
-
-                
+        no_augmentation_sources: Optional. List of sources to exclude for
+            augmentation. A source is string that identifies a dataset and is
+            defined in the Dataset class.
         """
         assert self.mode == "training", "Create model in training mode."
 
@@ -2339,8 +2340,7 @@ class MaskRCNN():
                                          batch_size=self.config.BATCH_SIZE,
                                          no_augmentation_sources=no_augmentation_sources)
         val_generator = data_generator(val_dataset, self.config, shuffle=True,
-                                       batch_size=self.config.BATCH_SIZE,
-                                       no_augmentation_sources=no_augmentation_sources)
+                                       batch_size=self.config.BATCH_SIZE)
 
         # Callbacks
         callbacks = [
@@ -2351,7 +2351,8 @@ class MaskRCNN():
         ]
 	
         # Add custom callbacks to the list
-        callbacks+=custom_callbacks
+        if custom_callbacks:
+            callbacks += custom_callbacks
 
         # Train
         log("\nStarting at epoch {}. LR={}\n".format(self.epoch, learning_rate))