[TMI'20] Unpaired Multi-modal Segmentation via Knowledge Distillation
107
stars
24
forks
source link
Solved The Tough Problem: tensorflow.python.framework.errors_impl.OutOfRangeError: RandomShuffleQueue '_1_shuffle_batch/random_shuffle_queue' is closed and has insufficient elements (requested 8, current size 0) #9
The code in main_combine.py need us to specify the path to the list of filenames: Medical-Cross-Modality-Domain-Adaptation/lists/mr_train_list, there are paths to all .tfrecords files in mr_train_list.
The problem is the code in train_combine.py reads those paths directly to find .tfrecords files on your machine, so
You need to modify the paths to the correct one using python's list derivation like in train_combine.py:
self.source_train_list = [os.path.join("~/Medical-Cross-Modality-Domain-Adaptation/lists", p) for p in source_train_list]
The code in
main_combine.pyneed us to specify the path to the list of filenames:Medical-Cross-Modality-Domain-Adaptation/lists/mr_train_list, there are paths to all.tfrecordsfiles inmr_train_list.The problem is the code in
train_combine.pyreads those paths directly to find.tfrecordsfiles on your machine, soYou need to modify the paths to the correct one using python's list derivation like in
train_combine.py: