DOC: Adds Inigo's week 8&9 blogposts #55
Conversation
…dicated folder 3 images from past posts had not been moved to the dedicated folder, so moved them
|
🪓 PR closed, deleted preview at https://github.com/dipy/preview-html/tree/main/dipy.org/pull/55/ |
itellaetxe
requested review from
jhlegarreta,
skoudoro,
pjsjongsung,
WassCodeur,
robinroy03 and
deka27
There was a problem hiding this comment.
Hi @itellaetxe,
Very interesting blog post. I learn new stuff so well explained.
See below a small comment.
also, we sure to add more spacing to make it easier to read. I know you already made a effort but I think you can add more paragraph and let the text breathe.
see your blog post preview here: https://dipy.github.io/preview-html/dipy.org/pull/55/posts/2024/2024_07_26_Inigo_week_9.html. or this one: https://dipy.github.io/preview-html/dipy.org/pull/55/posts/2024/2024_07_19_Inigo_week_8.html
Thank you in advance for updating that also
|
|
||
| This week was a bit shorter than usual because Thursday was a holiday in the Basque Country, and today we had an outdoor activity with my lab mates (we went kayaking to the Urdaibai Biosphere Reserve). Nevertheless, it was full of advances and interesting scientific matters. | ||
|
|
||
| As I mentioned in my last blog post, this week I worked on the implementation of the conditional VAE based on `this implementation <https://github.com/QingyuZhao/VAE-for-Regression/tree/master>` of this `paper <https://doi.org/10.1007/978-3-030-32245-8_91>`. |
There was a problem hiding this comment.
link does not work. wrong formatting I suppose. see here: https://dipy.github.io/preview-html/dipy.org/pull/55/posts/2024/2024_07_26_Inigo_week_9.html
great catch! thanks ! |
|
@skoudoro, I am getting this error while building locally: On the other hand, the CI fails but I am not familiar with this error: I think it has to do with how I have barely used Makefiles, so I don't know if this a silly question 😅 |
| :width: 600 | ||
|
|
||
| I also looked at the theoretical and technical implications of implementing the `beta-VAE architecture <https://openreview.net/forum?id=Sy2fzU9gl>`_ for my experiments, which could help in disentangling the latent space representation of the streamlines according to features learnt in an | ||
| unsupervised manner. |
There was a problem hiding this comment.
Unnecessary line break
| Shortly, applying a weight (bigger than 1) to the KL loss component of the VAE loss function encourages the model to learn a version of the latent space where features that can be perceived in the data space are aligned with the latent space dimensions. This way, one can modulate the generative process according to the learnt 'perceivable' features, once they are identified and located in the latent space. | ||
|
|
||
| However, increasing the beta weight compromises the reconstruction quality, which is what basically makes streamlines look reasonable. Finding a good beta weight is as 'simple' as running a hyperparameter search while constraining the parameter to be higher than one, and to try to prioritize | ||
| the MSE (Mean Squared Error, reconstruction loss) in the search algorithm. |
There was a problem hiding this comment.
Unnecessary line break
| :width: 600 | ||
|
|
||
| - Visual checking fiber generation for specific bundles. Knowing that different bundles have different fiber lengths, we try to generate fibers of specific length, and see whether the generated fibers belong to the desired bundle (no matter if they are plausible or implausible). | ||
| Having length as the conditioning variable allows us to perform this trick, what would not be so intuitive to check if we had used Fractional Anisotropy or other DTI-derived metrics, as these are not visually as intuitive as length. |
There was a problem hiding this comment.
Not sure how these two paragraphs are rendered.
There was a problem hiding this comment.
It is rendered as a full stop, no line breaks. I put it there for making text edition easier
|
Great Post. I need to learn more to understand the subject matter well. |
No idea, maybe
interesting error... I will investigate later today. Sorry, it has a lower priority in my todolist. Maybe @deka27 can look at it? check pydata_sphinx_theme requirement, and if we use a newer version. maybe we need to pin the version |
| :alt: First reconstruction of the training data of the conditional VAE (cVAE). | ||
| :width: 600 | ||
|
|
||
| On the other hand, note that the FiberCup has 7 distinct bundles, and both latent spaces show (when 2D-projected with the t-SNE algorithm) 7 clusters, suggesting that the network does know about the different bundles. Samples/streamlines are colored according to their length, and even if the bundle to which they belong is not plotted, we know that each cluster is formed by streamlines of different bundles because each bundle has a distinctive length. |
There was a problem hiding this comment.
I have been thinking about this lately, FURY will be a great fit for t-SNE viz. Should make a demo/function. Not sure yet.
There was a problem hiding this comment.
What are your ideas on that? Like implementing an interactive visualization for t-SNE?
Squashed 14 commits after rebasing on master to fix building error that Kaustav fixed.
itellaetxe
force-pushed
the
inigo_week_8_9_blogpost
branch
from
4bbeb3b to
cb83457
Compare
DOC: Adds Inigo's week 8&9 blogposts e16d3df
Also and adapts image paths to new dedicated folder (thanks @deka27 for creating it)
3 images from past posts had not been moved to the dedicated folder, so I moved them.