update

lanenet_model/lanenet.py

@@ -28,10 +28,10 @@ def __init__(self, phase, cfg):
         self._net_flag = self._cfg.MODEL.FRONT_END
 
         self._frontend = lanenet_front_end.LaneNetFrondEnd(
-            phase=phase, net_flag=self._net_flag
+            phase=phase, net_flag=self._net_flag, cfg=self._cfg
         )
         self._backend = lanenet_back_end.LaneNetBackEnd(
-            phase=phase
+            phase=phase, cfg=self._cfg
         )
 
     def inference(self, input_tensor, name, reuse=False):

lanenet_model/lanenet_back_end.py

@@ -10,29 +10,27 @@
 """
 import tensorflow as tf
 
-from local_utils.config_utils import parse_config_utils
 from lanenet_model import lanenet_discriminative_loss
 from semantic_segmentation_zoo import cnn_basenet
 
-CFG = parse_config_utils.lanenet_cfg
-
 
 class LaneNetBackEnd(cnn_basenet.CNNBaseModel):
     """
     LaneNet backend branch which is mainly used for binary and instance segmentation loss calculation
     """
-    def __init__(self, phase):
+    def __init__(self, phase, cfg):
         """
         init lanenet backend
         :param phase: train or test
         """
         super(LaneNetBackEnd, self).__init__()
+        self._cfg = cfg
         self._phase = phase
         self._is_training = self._is_net_for_training()
 
-        self._class_nums = CFG.DATASET.NUM_CLASSES
-        self._embedding_dims = CFG.MODEL.EMBEDDING_FEATS_DIMS
-        self._binary_loss_type = CFG.SOLVER.LOSS_TYPE
+        self._class_nums = self._cfg.DATASET.NUM_CLASSES
+        self._embedding_dims = self._cfg.MODEL.EMBEDDING_FEATS_DIMS
+        self._binary_loss_type = self._cfg.SOLVER.LOSS_TYPE
 
     def _is_net_for_training(self):
         """

tools/test_lanenet.py

@@ -86,7 +86,7 @@ def test_lanenet(image_path, weights_path):
 
     input_tensor = tf.placeholder(dtype=tf.float32, shape=[1, 256, 512, 3], name='input_tensor')
 
-    net = lanenet.LaneNet(phase='test')
+    net = lanenet.LaneNet(phase='test', cfg=CFG)
     binary_seg_ret, instance_seg_ret = net.inference(input_tensor=input_tensor, name='LaneNet')
 
     postprocessor = lanenet_postprocess.LaneNetPostProcessor()