[Enhancement] Val Dataset Transform Decoupled

contrib/Matting/bg_replace.py

@@ -20,10 +20,12 @@
 import paddle
 from paddleseg.cvlibs import manager, Config
 from paddleseg.utils import get_sys_env, logger
+from paddleseg.transforms import Compose
 
 from core import predict
 import model
 from dataset import MattingDataset
+from transforms import Compose
 from utils import get_image_list, estimate_foreground_ml
 
 
@@ -81,23 +83,14 @@ def main(args):
         raise RuntimeError('No configuration file specified.')
 
     cfg = Config(args.cfg)
-    val_dataset = cfg.val_dataset
-    if val_dataset is None:
-        raise RuntimeError(
-            'The verification dataset is not specified in the configuration file.'
-        )
-    elif len(val_dataset) == 0:
-        raise ValueError(
-            'The length of val_dataset is 0. Please check if your dataset is valid'
-        )
 
     msg = '\n---------------Config Information---------------\n'
     msg += str(cfg)
     msg += '------------------------------------------------'
     logger.info(msg)
 
     model = cfg.model
-    transforms = val_dataset.transforms
+    transforms = Compose(cfg.val_transforms)
 
     alpha = predict(
         model,

contrib/Matting/predict.py

@@ -21,6 +21,7 @@
 from core import predict
 from model import *
 from dataset import MattingDataset
+from transforms import Compose
 from utils import get_image_list
 
 
@@ -70,23 +71,14 @@ def main(args):
         raise RuntimeError('No configuration file specified.')
 
     cfg = Config(args.cfg)
-    val_dataset = cfg.val_dataset
-    if val_dataset is None:
-        raise RuntimeError(
-            'The verification dataset is not specified in the configuration file.'
-        )
-    elif len(val_dataset) == 0:
-        raise ValueError(
-            'The length of val_dataset is 0. Please check if your dataset is valid'
-        )
 
     msg = '\n---------------Config Information---------------\n'
     msg += str(cfg)
     msg += '------------------------------------------------'
     logger.info(msg)
 
     model = cfg.model
-    transforms = val_dataset.transforms
+    transforms = Compose(cfg.val_transforms)
 
     image_list, image_dir = get_image_list(args.image_path)
     if args.trimap_path is None:

paddleseg/cvlibs/config.py

@@ -300,16 +300,19 @@ def model(self) -> paddle.nn.Layer:
             raise RuntimeError('No model specified in the configuration file.')
         if not 'num_classes' in model_cfg:
             num_classes = None
-            if self.train_dataset_config:
-                if hasattr(self.train_dataset_class, 'NUM_CLASSES'):
-                    num_classes = self.train_dataset_class.NUM_CLASSES
-                elif hasattr(self.train_dataset, 'num_classes'):
-                    num_classes = self.train_dataset.num_classes
-            elif self.val_dataset_config:
-                if hasattr(self.val_dataset_class, 'NUM_CLASSES'):
-                    num_classes = self.val_dataset_class.NUM_CLASSES
-                elif hasattr(self.val_dataset, 'num_classes'):
-                    num_classes = self.val_dataset.num_classes
+            try:
+                if self.train_dataset_config:
+                    if hasattr(self.train_dataset_class, 'NUM_CLASSES'):
+                        num_classes = self.train_dataset_class.NUM_CLASSES
+                    elif hasattr(self.train_dataset, 'num_classes'):
+                        num_classes = self.train_dataset.num_classes
+                elif self.val_dataset_config:
+                    if hasattr(self.val_dataset_class, 'NUM_CLASSES'):
+                        num_classes = self.val_dataset_class.NUM_CLASSES
+                    elif hasattr(self.val_dataset, 'num_classes'):
+                        num_classes = self.val_dataset.num_classes
+            except FileNotFoundError:
+                pass
 
             if num_classes is not None:
                 model_cfg['num_classes'] = num_classes
@@ -402,3 +405,15 @@ def _is_meta_type(self, item: Any) -> bool:
 
     def __str__(self) -> str:
         return yaml.dump(self.dic)
+
+    @property
+    def val_transforms(self) -> list:
+        """Get val_transform from val_dataset"""
+        _val_dataset = self.val_dataset_config
+        if not _val_dataset:
+            return []
+        _transforms = _val_dataset.get('transforms', [])
+        transforms = []
+        for i in _transforms:
+            transforms.append(self._load_object(i))
+        return transforms

predict.py

@@ -18,8 +18,9 @@
 import paddle
 
 from paddleseg.cvlibs import manager, Config
-from paddleseg.utils import get_sys_env, logger, config_check, get_image_list
+from paddleseg.utils import get_sys_env, logger, get_image_list
 from paddleseg.core import predict
+from paddleseg.transforms import Compose
 
 
 def parse_args():
@@ -141,24 +142,18 @@ def main(args):
         raise RuntimeError('No configuration file specified.')
 
     cfg = Config(args.cfg)
-    val_dataset = cfg.val_dataset
-    if not val_dataset:
-        raise RuntimeError(
-            'The verification dataset is not specified in the configuration file.'
-        )
 
     msg = '\n---------------Config Information---------------\n'
     msg += str(cfg)
     msg += '------------------------------------------------'
     logger.info(msg)
 
     model = cfg.model
-    transforms = val_dataset.transforms
+    transforms = Compose(cfg.val_transforms)
     image_list, image_dir = get_image_list(args.image_path)
     logger.info('Number of predict images = {}'.format(len(image_list)))
 
     test_config = get_test_config(cfg, args)
-    config_check(cfg, val_dataset=val_dataset)
 
     predict(
         model,