New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
[skip travis] EG: add example notebook on advanced model techniques #174
Conversation
Check out this pull request on See visual diffs & provide feedback on Jupyter Notebooks. Powered by ReviewNB |
adeb10f
to
7d0cd2b
Compare
View / edit / reply to this conversation on ReviewNB DavidBAEpstein commented on 2021-11-05T21:29:23Z Instead of "pass" above, it might be clearer to write "# your code here" should be inherited from -> should inherit from be of the following structure -> have the following structure which has the above structure -> which have the above structure here we define -> we define here the same way how a semantic -> the same way that a semantic kernels on the input image -> kernels to the input image it can be directly fit in -> it can be directly fitted (your usage may possibly be US English---not sure about this) mostafajahanifar commented on 2021-11-08T18:09:21Z I added "# your code here" expression but kept the pass as well, because when there is no code, they should have the "pass" statement at the end.
|
View / edit / reply to this conversation on ReviewNB DavidBAEpstein commented on 2021-11-05T21:29:23Z In TIAToolbox, configurations as we have done -> configurations, as we have done cell bellow (for more information on -> cell below. (For more information on |
View / edit / reply to this conversation on ReviewNB DavidBAEpstein commented on 2021-11-05T21:29:24Z through your own pretrain classification -> your own pretrained classification there are some times that you -> sometimes you in TIAToolbox framework -> in the TIAToolbox framework an external pretrained -> an externally pretrained that is not included -> not included so it would be suitable for TIAToolbox pipeline -> so that it would be suitable for a TIAToolbox pipeline model class should follow the model class should follow the vqdang commented on 2021-11-05T21:58:58Z In the example bolow > In the example below |
View / edit / reply to this conversation on ReviewNB DavidBAEpstein commented on 2021-11-05T21:29:25Z knowledge from -> knowledge of The functionality of these functions should be designed based - The design of these functions should be based TIAToolbox UNet architecutre -> TIAToolbox UNet architecture get a clue > get an idea Anyway, now that -> Now that can be ultimately -> can ultimately be applied the applied the
|
View / edit / reply to this conversation on ReviewNB DavidBAEpstein commented on 2021-11-05T21:29:25Z has been predicted rightfully > has been correctly predicted |
View / edit / reply to this conversation on ReviewNB DavidBAEpstein commented on 2021-11-05T21:29:26Z the inference process of multiple WSIs -> the inference process for multiple WSIs such as provided input -> for example, provided input has been utilized -> is utilized supports (a list of supported WSI formats can be found here. -> supports. A list of supported WSI formats can be found here. all types of plane images as OpenCv does).-> the same types of plane images as OpenCV. related to related to you will risk -> you risk we dont modify -> we don't modify and change its source image for reading -> and uses this information to decide which source image to read is also being queried from the shared -> is also queried through the shared using Tiattolbox -> using TIAToolbox we present this example where -> we present the example below, where to allow loading a 'multiplex tif' tile, which is not officially supported by TIAToolbox -> to load a 'multiplex tif' tile, although this is not officially supported by TIAToolbox define the image reader function -> define an image reader function |
View / edit / reply to this conversation on ReviewNB DavidBAEpstein commented on 2021-11-05T21:29:27Z Line #9. # here, we create a virtual whole-slide image from the image array, but with a single level. here we create a virtual whole slide image, with a single level, from the image array |
View / edit / reply to this conversation on ReviewNB DavidBAEpstein commented on 2021-11-05T21:29:27Z Do you want to capitalize XReader? It's all lower case in the code. based Furthermore, we should make sure that We need to make sure that in this case, mostafajahanifar commented on 2021-11-08T18:48:52Z xreader is a function, on the other hand, XReader is class that uses xreader function. |
View / edit / reply to this conversation on ReviewNB DavidBAEpstein commented on 2021-11-05T21:29:28Z Now we have both -> Now that we have both can use the in the prediction framework -> can use the prediction framework section but we will overwrite the section.We will overwrite the (returning model input in the output allows us to check what model is seeing in the input) -> Returning model input as the output allows us to check what our model sees in the input. |
View / edit / reply to this conversation on ReviewNB DavidBAEpstein commented on 2021-11-05T21:29:28Z Everything is in place, let's instantiate our model, segmentor, -> Everything is in place. Let's instantiate our model segmentor, here we predict on -> here we predict from |
View / edit / reply to this conversation on ReviewNB DavidBAEpstein commented on 2021-11-05T21:29:29Z input and output image ->input and output images |
In the example bolow > In the example below View entire conversation on ReviewNB |
View / edit / reply to this conversation on ReviewNB vqdang commented on 2021-11-05T22:02:03Z external_model.infer_batch=infer_batch external_model.preproc_func=preproc_func external_model.postproc_func=postproc_func
^^^ Fix spacing at mostafajahanifar commented on 2021-11-08T18:21:04Z Sure, dear flake-8 :-p |
View / edit / reply to this conversation on ReviewNB vqdang commented on 2021-11-05T22:02:04Z
##2. Dealing with exotic data formats
^^^ vscode headline doesnt work here |
I added "# your code here" expression but kept the pass as well, because when there is no code, they should have the "pass" statement at the end.
View entire conversation on ReviewNB |
Sure, dear flake-8 :-p View entire conversation on ReviewNB |
xreader is a function, on the other hand, XReader is class that uses xreader function. View entire conversation on ReviewNB |
Codecov Report
@@ Coverage Diff @@
## develop #174 +/- ##
========================================
Coverage 99.74% 99.74%
========================================
Files 44 44
Lines 3540 3540
Branches 576 576
========================================
Hits 3531 3531
Misses 3 3
Partials 6 6 Continue to review full report at Codecov.
|
…example-advanced
Adding a new example notebook on semantic segmentation (more generally, model prediction) task where advanced topics are addressed. Particularly, in this notebook we are trying to demonestrate how user can use Tiatoolbox to solve the problems in the following three scenarios:
1. Instead of pretrained models embedded in Tiatoolbox's repsitory, user wants to use your own deep learning model (in Pytorch) in the Tiatoolbox prediction workflow.
2. User's input data is of an exotic form which the Tiatoolbox data stream functionality does not support by default.