# Descriptor dimensionality reduction

**URL:** <https://discourse.flucoma.org/t/descriptor-dimensionality-reduction/606>\
**Category:** Pre-Release Toolbox2 New Ideas\
**Created:** [July 24, 2020, 2:55pm UTC](https://discourse.flucoma.org/t/descriptor-dimensionality-reduction/606 "2020-07-24T14:55:14Z")\
**Posts on this page:** 3\
**Page:** 2

<div class="post-metadata">

**Author:** ![jamesbradbury](https://discourse.flucoma.org/user_avatar/discourse.flucoma.org/jamesbradbury/32/1252_2.png) [@jamesbradbury](https://discourse.flucoma.org/u/jamesbradbury)\
**Post date:** [August 2, 2020, 10:32am UTC](https://discourse.flucoma.org/t/descriptor-dimensionality-reduction/606/21 "2020-08-02T10:32:52Z")

</div>

In speculative funky land this is quite promising for python heads

> **[maxfrenzel/CompressionVAE](https://github.com/maxfrenzel/CompressionVAE)**
>
> General-purpose dimensionality reduction and manifold learning tool based on Variational Autoencoder, implemented in TensorFlow. - maxfrenzel/CompressionVAE

---

<div class="post-metadata">

**Author:** ![b.hackbarth](https://discourse.flucoma.org/letter_avatar_proxy/v4/letter/b/8e8cbc/32.png) [@b.hackbarth](https://discourse.flucoma.org/u/b.hackbarth)\
**Post date:** [August 2, 2020, 10:56am UTC](https://discourse.flucoma.org/t/descriptor-dimensionality-reduction/606/22 "2020-08-02T10:56:38Z")

</div>

> [@tedmoore](#):
>
> This makes a lot of sense to me. I think it’s a really nice lens for concatenative synthesis. I have a [plotter tool](https://github.com/tedmoore/PlotXYColor) in SC that I use a lot – I might edit that tool to make it possible to move points around and see what that sounds/feels like.

Thanks for the cool video. Great to be able to explore the different dimension reduction methods side by side. I think that the best thing for me to do in this instance is to make a video showing a bit of what I’m thinking about/trying with regard to visual target editing. Will try to get to it next week.

In the meantime, have you considered plotting a 4th dimension as the size of the points? I’ve been doing this with loudness and it’s pretty useful visually.

---

<div class="post-metadata">

**Author:** ![tremblap](https://discourse.flucoma.org/user_avatar/discourse.flucoma.org/tremblap/32/5_2.png) [@tremblap](https://discourse.flucoma.org/u/tremblap)\
**Post date:** [August 2, 2020, 11:55am UTC](https://discourse.flucoma.org/t/descriptor-dimensionality-reduction/606/23 "2020-08-02T11:55:10Z")

</div>

> [@b.hackbarth](#):
>
> Great to be able to explore the different dimension reduction methods side by side.

Indeed that was the fun part of [this project](https://discourse.flucoma.org/t/edit-no-more-confidential-sharing-experimental-results-on-descriptor-space-dimension-reduction-experiment/161) I was referring to on Friday. If you are ok in SC you can try it for yourself [here](https://github.com/flucoma/FluidCorpusMap) - the fun bit is that we all had a favourite that was different 🙂

[Previous page](https://discourse.flucoma.org/t/descriptor-dimensionality-reduction/606.md?page=1)