label normalisation layer requires label input

[mcbrs1a]

Hi,

I am running an inference using a model I built on a different machine. My config file sets label normalisation as true; the label and image file exists in it's directory, the correct location is output in the .csv file produced. However when running inference I obtain:

AssertionError: label normalisation layer requires label input, however it is not provided in the config file.

Any suggestions? Below is my config file:

[label] #path_to_search = C:/Users/Victor/niftynet/data/my_network_collection_dense_vnet path_to_search = /home/petic/niftynet/data/my_network_collection_dense_vnet/ filename_contains = Label #spatial_window_size = (100, 100, 300) #spatial_window_size = (72, 80, 352) spatial_window_size = (64, 48, 160) pixdim = 0.2512, 0.2512, 0.2512 #pixdim = 1, 1, 1 axcodes=(A, R, S) interp_order = 0

[ct] #path_to_search = C:/Users/Victor/niftynet/data/my_network_collection_dense_vnet path_to_search = /home/petic/niftynet/data/my_network_collection_dense_vnet/ filename_contains = CT

#spatial_window_size = (100, 100, 300) #96, 104, 288
#88, 96, 280
#80, 88, 272
#72, 80, 264
#64, 72, 256 #spatial_window_size = (72, 80, 352) spatial_window_size = (64, 48, 160) pixdim = 0.2512, 0.2512, 0.2512 #pixdim = 1, 1, 1 axcodes=(A, R, S) interp_order = 3

############################## system configuration sections [SYSTEM] cuda_devices = "" num_threads = 1 num_gpus = 0 #model_dir = C:/Users/Victor/niftynet/models/my_network_collection_dense_vnet model_dir = /home/petic/niftynet/models/my_network_collection_dense_vnet

[NETWORK] name = dense_vnet activation_function = prelu batch_size = 1 volume_padding_size = 0 #window_sampling = resize #queue_length = 36 #histogram_ref_file = histogramLT2T_9596.txt histogram_ref_file = histogramAll2T_9495.txt norm_type = percentile cutoff = (0.01, 0.99) normalisation = True whitening = True normalise_foreground_only=True foreground_type = otsu_plus multimod_foreground_type = and window_sampling = balanced #window_sampling = resize #window_sampling = uniform queue_length = 20

[TRAINING] sample_per_volume = 32 lr = 0.001 #loss_type = my_network_collection_dense_vnet.dice_hinge.dice loss_type = Dice starting_iter = 0 save_every_n = 100 max_iter = 3001 rotation_angle = (-10.0, 10.0) scaling_percentage = (-10.0, 10.0)

save_every_n / tensorboard_every_n was 10 / max_iter was 10001 / max_checkpoints was 101

[INFERENCE]

border = (0, 0, 0) inference_iter = 3000 #save_seg_dir = C:/Users/Victor/niftynet/models/my_network_collection_dense_vnet/segmentation_output/ save_seg_dir =/home/petic/niftynet/models/my_network_collection_dense_vnet/segmentation_output

#spatial_window_size = (192, 112, 328) spatial_window_size = (112, 104, 384) #spatial_window_size = (256, 256, 176) output_interp_order = 0

############################ custom configuration sections [SEGMENTATION] image = ct label = label sampler = label output_prob = False label_normalisation = True num_classes = 9 min_numb_labels = 9 min_sampling_ratio = 0.000001