add compose

nndet/io/transforms/__init__.py

-from nndet.io.transforms.base import AbstractTransform
+from nndet.io.transforms.base import (
+    AbstractTransform,
+    Compose,
+    )
 from nndet.io.transforms.instances import (
     Instances2Boxes,
     Instances2Segmentation,

nndet/io/transforms/base.py

-from typing import Any
+from typing import Any, Sequence
 
 import torch
 
@@ -30,4 +30,27 @@ class AbstractTransform(torch.nn.Module):
             context = torch.no_grad()
 
         with context:
-            return super().__call__(*args, **kwargs)
\ No newline at end of file
+            return super().__call__(*args, **kwargs)
+
+
+class Compose(AbstractTransform):
+    def __init__(self, *transforms):
+        """
+        Compose multiple transforms to one
+        
+        Args:
+            transforms: transformations to compose
+        """
+        super().__init__(grad=False)
+        if len(transforms) == 1 and isinstance(transforms[0], Sequence):
+            transforms = transforms[0]
+
+        self.transforms = torch.nn.ModuleList(list(transforms))
+
+    def forward(self, **batch):
+        """
+        Augment batch
+        """
+        for t in self.transforms:
+            batch = t(**batch)
+        return batch

nndet/ptmodule/retinaunet/base.py

@@ -62,8 +62,12 @@ from nndet.inference.helper import predict_dir
 from nndet.inference.ensembler.segmentation import SegmentationEnsembler
 from nndet.inference.ensembler.detection import BoxEnsemblerSelective
 
-from rising.transforms import Compose
-from nndet.io.transforms import Instances2Boxes, Instances2Segmentation, FindInstances
+from nndet.io.transforms import (
+    Compose,
+    Instances2Boxes,
+    Instances2Segmentation,
+    FindInstances,
+    )
 
 
 class RetinaUNetModule(LightningBaseModuleSWA):