Comments (6)
yyguo0536
commented on August 21, 2024
1
Hi there, please see the following,
- What type of file for input ? Is it 3D file, or 2D slices ?
- it's 3D file.
- What type of medical image can be used ? DICOM, nifti ?
- nifti file.
- What are the dimensions that are allowed ? Is 96 the maximum ?
- This is depended on your GPU memory. My setting is 128x128x96.
- What kind of value should be in the column 'training file' and 'ground_truth' ? and why 24, 48, 96 ?
- 'training file' is 4D intensity images, and 'ground_truth' is their corresponding segmentations which is used to evaluation.
from svin.
yyguo0536
commented on August 21, 2024
1
I got it. In line 45 of "net_3d.py", "self.convb = VNetConvBlock(in_channels, out_channels, layers=layers, kernel_sz=k_size, pad=pad_size)" should be "self.convb = NetConvBlock(in_channels, out_channels, layers=layers, kernel_sz=k_size, pad=pad_size)"
from svin.
cyiheng
commented on August 21, 2024
Thank you for you answer !
For the csv, I should have ask for the motion-net first about the "image" and "annotation" column, what are they suppose to refer ? The same as for "training_file" and "ground_truth" ?
data = pd.read_csv('your training data with scale/4')
training_l = data['image']
label_l = data['annotation']
for i in range(len(training_l)):
training_rows24.append(training_l[i])
trainning_label24.append(label_l[i])
And I tried for the motion_net and I get this error actually, maybe you can tell me more about what it expect as input ?
I printed the shape of the combinedimage (first line)
('combinedimage 12', (1, 2, 96, 96, 96))
Traceback (most recent call last):
File "train_motion.py", line 115, in <module>
model.optimize_parameters()
File "/home/yiheng/bacasable/SVIN/motion-net/models/motion_model.py", line 250, in optimize_parameters
self.forward()
File "/home/yiheng/bacasable/SVIN/motion-net/models/motion_model.py", line 169, in forward
self.real96_A2, self.real48_A2, self.real24_A2)
File "/home/yiheng/anaconda3/envs/svin/lib/python2.7/site-packages/torch/nn/modules/module.py", line 532, in __call__
result = self.forward(*input, **kwargs)
File "/home/yiheng/anaconda3/envs/svin/lib/python2.7/site-packages/torch/nn/parallel/data_parallel.py", line 150, in forward
return self.module(*inputs[0], **kwargs[0])
File "/home/yiheng/anaconda3/envs/svin/lib/python2.7/site-packages/torch/nn/modules/module.py", line 532, in __call__
result = self.forward(*input, **kwargs)
File "/home/yiheng/bacasable/SVIN/motion-net/models/net_3d.py", line 263, in forward
field12_24, field12_48, field12_96 = self.forward_branch(combined12)
File "/home/yiheng/bacasable/SVIN/motion-net/models/net_3d.py", line 235, in forward_branch
br1 = self.in_block(combined_image)
File "/home/yiheng/anaconda3/envs/svin/lib/python2.7/site-packages/torch/nn/modules/module.py", line 532, in __call__
result = self.forward(*input, **kwargs)
File "/home/yiheng/bacasable/SVIN/motion-net/models/net_3d.py", line 51, in forward
out = self.convb(x)
File "/home/yiheng/anaconda3/envs/svin/lib/python2.7/site-packages/torch/nn/modules/module.py", line 532, in __call__
result = self.forward(*input, **kwargs)
File "/home/yiheng/bacasable/SVIN/motion-net/models/vnet.py", line 34, in forward
out = self.convs[i](out)
File "/home/yiheng/anaconda3/envs/svin/lib/python2.7/site-packages/torch/nn/modules/module.py", line 532, in __call__
result = self.forward(*input, **kwargs)
File "/home/yiheng/anaconda3/envs/svin/lib/python2.7/site-packages/torch/nn/modules/conv.py", line 345, in forward
return self.conv2d_forward(input, self.weight)
File "/home/yiheng/anaconda3/envs/svin/lib/python2.7/site-packages/torch/nn/modules/conv.py", line 342, in conv2d_forward
self.padding, self.dilation, self.groups)
RuntimeError: Expected 4-dimensional input for 4-dimensional weight 24 2 3 3, but got 5-dimensional input of size [1, 2, 96, 96, 96] instead
Thank you in advance again !
from svin.
yyguo0536
commented on August 21, 2024
Hi, the 'annotation' is the same as the 'groundtruth' for the evaluation.
For your error, can you set "which_model_netG" to "motion"?
from svin.
cyiheng
commented on August 21, 2024
It was already set to "motion".
I used the line on the main screen of the github.
Here the options below
----------------- Options ---------------
batchSize: 1
beta1: 0.5
border_ratio: 0.0277777777778
checkpoints_dir: ./save_motion/ [default: ./checkpoints]
classes: [0, 1]
continue_train: False
crop: None
dataroot: ./ [default: None]
dataset_mode: aligned
display_freq: 200
display_id: 0 [default: 1]
display_ncols: 4
display_port: 8097
display_server: http://localhost
display_winsize: 256
epoch_count: 1
evaluation: False
fineSize: 256
gpu_ids: 0
init_gain: 0.02
init_type: normal
input_nc: 1
isTrain: True [default: None]
lambda_L1: 5000.0
loadSize: 286
lr: 0.0001
lr_decay_iters: 50
lr_policy: lambda
max_dataset_size: inf
model: motion
nThreads: 4
n_layers_D: 3
name: experiment_name
ndf: 64
net3d_dir_D: ./results/
net3d_dir_G: ./results/
ngf: 64
niter: 1000
niter_decay: 1000
no_dropout: False
no_flip: False
no_html: False
no_lsgan: True
norm: batch
output_nc: 1
phase: train
pool_size: 0
print_freq: 1
resize_or_crop: resize_and_crop
rotate: 0.0872664625997
save_epoch_freq: 10
save_latest_freq: 5000
serial_batches: False
spacing: None
suffix:
update_html_freq: 1000
verbose: False
which_direction: AtoB
which_epoch: latest
which_model_netD: basic
which_model_netG: motion [default: unet_256]
----------------- End -------------------
from svin.
cyiheng
commented on August 21, 2024
Yes ! Thank you very much !
It works now !
from svin.
Related Issues (9)
- from .warp_layer import Dense3DSpatialTransformer ImportError: cannot import name Dense3DSpatialTransformer HOT 2
- File "/media/sc1/BED0CCE0D0CCA04F/lizhipeng/BIS/SVIN/interpolation/motion_model.py", line 4, in <module> from grid import Dense3DSpatialTransformer ImportError: No module named grid HOT 1
- Some question about the idea HOT 1
- Questions about Senquential Volumetric Interpolation HOT 2
- Out of memory HOT 1
- Testing HOT 2
- Dataset
- SS-BL-Net
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google ❤️ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from svin.