i want to train a neural network with flexible output size. In the beginning, i used the matlab deep network designer to manually replace the classification and fully connected layer to the desired output size. Now, i want to automatically replace it, using a script.
Which command is working for that?
Simply trying the line:
net.Layers(142,1).InputSize = 10;
gives me the error message
Unable to set the 'InputSize' property of class 'FullyConnectedLayer' because it is read-only.
Trying to replace the complete layer (not only inputsize) is resulting in the same error message.
Is this possible with matlab, and if yes, which commands will do the job?
Thanks in advance!
Okay, i think i solved it. When changing the network manually inside of the deepnetworkdesigner, there is the possibility to click "export code" which exports the desired code for the adapted output size
Related
I want to segment pelvic in MRI from the SMIR dataset using the MONAILABEL plugin. I have read relatively a lot about this plugin. However, I can’t perform the segmentation well enough yet.
These are the steps I take to do so:
connecting the server using the anaconda prompt
enabling the plugin in Slicer and loading one image
labeling the pelvis manually using the paint button in the scribbles part and the updating
click “submit label”
and then repeating the process for some other images while the network is being trained.
After the training process, I actually could not see anything when opening the mask files.
I also encounter these two errors:
" AssertionError: Not a valid Label "
" TypeError: object of type ‘NoneType’ has no len() "
Is it the correct way of using MONAI or am I missing something?
Which aspects should I take into account before starting the process?
I would appreciate your help and suggestions.
Thank you
I've been following the Build your own image classifier using Transfer Learning tutorial found at this link.
At the end of the tutorial, it says you use your own data set, which I created by following the example in the tutorial. It successfully completed and created a folder ~/CNTK-Samples-2-3-1/Examples/Image/TransferLearning/Output, which is expected. Inside this output folder is predictions.txt, predOutput.txt, and TransferLearning.model.
My question is, how do I access TransferLearning.model? I'm not sure what to do with it and I can't find anything in the documentation allowing me to alter it or run it.
Once you successfully create your model, you can use it in Python environment (with CNTK) to classify your images. The code should look something like this:
#load the trained transfer learning model
trained_model = load_model(model_file)
#for every new image:
#get predictions for a single image
probs = eval_single_image(trained_model, img_file, image_width, image_height)
I've been trying to use raster calculation in ipython for a tif file I have uploaded, but I'm unable to find the whole code for the function. I keep finding examples such as below, but am unsure how to use this.
gdal_calc.py -A input.tif --outfile=result.tif --calc="A*(A>0)" --NoDataValue=0
I then tried another process by assigning sections, however this still doesn't work (code below)
a = '/iPythonData/cstone/prec_7.tif'
outfile = '/iPythonData/cstone/prec_result.tif'
expr = 'A<125'
gdal_calc.py -A=a --outfile=outfile --calc='expr' --NoDataValue=0
It keeps coming up with can't assign to operator. Can someone please help with the whole code.
Looking at the source code for gdal_calc.py, the file is only about 300 lines. Here is a link to that file.
https://raw.githubusercontent.com/OSGeo/gdal/trunk/gdal/swig/python/scripts/gdal_calc.py
The punchline is that they just create an OptionParser object in main and pass it to the doit() method (Line 63). You could generate the same OptionParser instance based on the same arguments you pass to it via the command-line and call their doit method directly.
That said, a system call is perfectly valid per #thomas-k. This is only if you really want to stay in the Python environment.
I am working on an experiment that take a lot of data samples and save it to different video files. For example, I take data and save it to a file called "slowmotion.avi" and then I take another set of data and save it to another file called "normalspeed.avi".
I am trying to find a way that I could change the name of file being saved without editing the code. The way I am using right now makes me have to open the code and change the name of the file directory within the code and then save the code. I want to make it a lot faster and easier to use.
I have tried the following code to fix this problem but it doesn't work.
graph=input('Graph of experiment: ');
vidObj = VideoWriter('%3.1f.avi\n',graph);
.....
Hope I didn't confuse you.
A possible solution:
graph=input('Graph of experiment: ','s');
vidObj = VideoWriter([graph,'.avi']);
The 's' in the input() function indicates that the expected input is a string, and [graph,'.avi'] simply concatenates both strings.
I have got this really strange error in matlab. When I try to run the command
save(fullfile('filepath','filename'),'var','-v7');
I get the error message,
error using save can't write file
but when I try
save(fullfile('filepath','filename'),'var','-v7.3');
everything works fine. The the variable takes some space on the workspace, 165MB, but the I would guess that the size should not be an issue here. Does anyone know why it does not work to save in v7?
For the one that want to confirm the size of the variable, I will add the whos information,
Name Size Bytes Class Attributes
myName 1x1 173081921 struct
BR/ Patrik
EDIT
The variable I try to save is a struct with plenty of fields. I have tried to save a 3 dimensional matrix of size 800 mb, which went through without problems.
This is not an exact match to your problem, but I received the same error message when trying to save to -v6 format. Matlab is supposed to issue an error when a variable type or size is not supported:
help save
...
If any data items require features that the specified version does not support, MATLAB does not save those items and issues a warning. You cannot specify a version later than your version of MATLAB software.
Matlab's error checking seems to not be perfect, because there are certain situations (dependent on Matlab version and the particular variable type) that just fail all together with this not so helpful error message:
Error using save
Can't write file filename.mat.
For example, saving a string with certain unicode characters with the '-v6' option in Matlab r2015b in Linux produces the error, but Matlab r2016a in Windows does not. This is the output from my Matlab r2015b session:
>> A=char(double(65533))
A =
?
>> save('filename.mat','-v6','A')
Error using save
Can't write file filename.mat.
Without having your specific variable to test with, but seeing that the error messages match, I suggest removing parts of your data structure until it will save in order to isolate the variable that is causing it to fail.