cfg_net.txt

[network1] # for detection
# type of network
net_type = PNet

# name of network
net_name = pnet

# input patch shape e.g, [D, H, W, C]
data_shape = [1, 96, 96, 1]

# output patch shape, [D, H, W, C]
label_shape = [1, 96, 96, 1]

# number of class
class_num = 2

# model_file
model_file = model/detect/pnet.ckpt

[network1_parameter]

[network2] # for segmentation
# type of network
net_type = PNet

# name of network
net_name = pnet_seg

# input patch shape e.g, [D, H, W, C]
data_shape = [1, 96, 96, 1]

# output patch shape, [D, H, W, C]
label_shape = [1, 96, 96, 1]

# number of class
class_num = 2

# model_file
model_file = model/segment/pnet_seg.ckpt

[network2_parameter]

[detect_testing]
fix_batch_size = False

use_depth_as_batch_size = True

# max batch size
batch_size = 80

resize_input = True
resize_input_to_given_shape_in_2d = [96, 96]

# 1, using fixed tensor shape (default), 2, adaptive censor shape in 2D, 3, adaptive censorship in 3D
shape_mode = 1

[segment_testing]
fix_batch_size = False

use_depth_as_batch_size = True

# max batch size
batch_size = 40

resize_input = True
resize_input_to_given_shape_in_2d = [96, 96]

# 1, using fixed tensor shape (default), 2, adaptive censor shape in 2D, 3, adaptive censorship in 3D
shape_mode = 1