[Fix] Fix magnitude_range in RandAug

mmcls/datasets/pipelines/auto_augment.py

@@ -67,8 +67,9 @@ class RandAugment(object):
             augmentation. For those which have magnitude, (given to the fact
             they are named differently in different augmentation, )
             `magnitude_key` and `magnitude_range` shall be the magnitude
-            argument (str) and the range of magnitude (tuple in the format or
-            (minval, maxval)), respectively.
+            argument (str) and the range of magnitude (tuple in the format of
+            (val1, val2)), respectively. Note that val1 is not necessarily
+            less than val2.
         num_policies (int): Number of policies to select from policies each
             time.
         magnitude_level (int | float): Magnitude level for all the augmentation
@@ -85,6 +86,10 @@ class RandAugment(object):
     Note:
         `magnitude_std` will introduce some randomness to policy, modified by
         https://github.com/rwightman/pytorch-image-models
+        When magnitude_std=0, we calculate the magnitude as follows:
+
+        .. math::
+            magnitude = magnitude_level / total_level * (val2 - val1) + val1
     """
 
     def __init__(self,
@@ -130,18 +135,20 @@ def _process_policies(self, policies):
             processed_policy = copy.deepcopy(policy)
             magnitude_key = processed_policy.pop('magnitude_key', None)
             if magnitude_key is not None:
-                minval, maxval = processed_policy.pop('magnitude_range')
+                val1, val2 = processed_policy.pop('magnitude_range')
                 magnitude_value = (self.magnitude_level / self.total_level
-                                   ) * float(maxval - minval) + minval
+                                   ) * float(val2 - val1) + val1
 
                 # if magnitude_std is positive number or 'inf', move
                 # magnitude_value randomly.
+                maxval = max(val1, val2)
+                minval = min(val1, val2)
                 if self.magnitude_std == 'inf':
                     magnitude_value = random.uniform(minval, magnitude_value)
                 elif self.magnitude_std > 0:
                     magnitude_value = random.gauss(magnitude_value,
                                                    self.magnitude_std)
-                    magnitude_value = min(maxval, max(0, magnitude_value))
+                    magnitude_value = min(maxval, max(minval, magnitude_value))
                 processed_policy.update({magnitude_key: magnitude_value})
             processed_policies.append(processed_policy)
         return processed_policies

tests/test_pipelines/test_auto_augment.py

@@ -185,6 +185,34 @@ def test_rand_augment():
     # apply rotation with prob=0.
     assert (results['img'] == results['ori_img']).all()
 
+    # test case where magnitude_range is reversed
+    random.seed(1)
+    np.random.seed(0)
+    results = construct_toy_data()
+    reversed_policies = [
+        dict(
+            type='Translate',
+            magnitude_key='magnitude',
+            magnitude_range=(1, 0),
+            pad_val=128,
+            prob=1.,
+            direction='horizontal'),
+        dict(type='Invert', prob=1.),
+        dict(
+            type='Rotate',
+            magnitude_key='angle',
+            magnitude_range=(30, 0),
+            prob=0.)
+    ]
+    transform = dict(
+        type='RandAugment',
+        policies=reversed_policies,
+        num_policies=1,
+        magnitude_level=30)
+    pipeline = build_from_cfg(transform, PIPELINES)
+    results = pipeline(results)
+    assert (results['img'] == results['ori_img']).all()
+
     # test case where num_policies = 2
     random.seed(0)
     np.random.seed(0)