More updates in the ribbon segmentation

Author: constantinpape

scripts/inner_ear/training/postprocessing_and_evaluation.py

@@ -13,8 +13,8 @@
 
 from train_structure_segmentation import get_train_val_test_split
 
-ROOT = "/home/pape/Work/data/synaptic_reconstruction/moser"
-# ROOT = "/mnt/lustre-emmy-hdd/projects/nim00007/data/synaptic-reconstruction/moser"
+# ROOT = "/home/pape/Work/data/synaptic_reconstruction/moser"
+ROOT = "/mnt/lustre-emmy-hdd/projects/nim00007/data/synaptic-reconstruction/moser"
 MODEL_PATH = "/mnt/lustre-emmy-hdd/projects/nim00007/models/synaptic-reconstruction/vesicle-DA-inner_ear-v2"
 OUTPUT_ROOT = "./predictions"
 

scripts/inner_ear/training/structure_prediction_and_evaluation.py

@@ -11,8 +11,8 @@
 from train_structure_segmentation import get_train_val_test_split
 from train_structure_segmentation import noop  # noqa
 
-# ROOT = "/mnt/lustre-emmy-hdd/projects/nim00007/data/synaptic-reconstruction/moser"
-ROOT = "/home/pape/Work/data/synaptic_reconstruction/moser"
+ROOT = "/mnt/lustre-emmy-hdd/projects/nim00007/data/synaptic-reconstruction/moser"
+# ROOT = "/home/pape/Work/data/synaptic_reconstruction/moser"
 OUTPUT_ROOT = "./predictions"
 
 
@@ -176,13 +176,13 @@ def predict_and_evaluate_target_domain(paths, name, adapted_model_path):
 
 def predict_and_evaluate_vesicle_pools():
     paths = sorted(glob(os.path.join(ROOT, "other_tomograms/01_vesicle_pools", "*.h5")))
-    adapted_model_path = "./checkpoints/structure-model-adapt-vesicle_pools"
+    adapted_model_path = "./checkpoints/structure-model-adapt-vesicle_pools-v2"
     predict_and_evaluate_target_domain(paths, "vesicle_pools", adapted_model_path)
 
 
 def predict_and_evaluate_rat():
     paths = sorted(glob(os.path.join(ROOT, "other_tomograms/03_ratten_tomos", "*.h5")))
-    adapted_model_path = "./checkpoints/structure-model-adapt-rat"
+    adapted_model_path = "./checkpoints/structure-model-adapt-rat-v2"
     predict_and_evaluate_target_domain(paths, "rat", adapted_model_path)
 
 

synaptic_reconstruction/inference/ribbon_synapse.py

@@ -36,7 +36,7 @@ def segment_ribbon_synapse_structures(
         and the predictions if return_predictions is True.
     """
     if verbose:
-        print("Segmenting mitochondria in volume of shape", input_volume.shape)
+        print("Segmenting ribbon synapse structures in volume of shape", input_volume.shape)
     # Create the scaler to handle prediction with a different scaling factor.
     scaler = _Scaler(scale, verbose)
     input_volume = scaler.scale_input(input_volume)
@@ -49,7 +49,7 @@ def segment_ribbon_synapse_structures(
     assert len(structure_names) == predictions.shape[0]
 
     predictions = {
-        name: scaler.scale_output(predictions[i], is_segmentation=False) for i, name in enumerate(structure_names)
+        name: scaler.rescale_output(predictions[i], is_segmentation=False) for i, name in enumerate(structure_names)
     }
     if threshold is not None:
         for name in structure_names: