Also it's open source:
🔗 https://github.com/kirel/detexify
and there is sample data to train from for #AI and #ML researchers:
🔗 https://github.com/kirel/detexify-data
Also it's open source:
🔗 https://github.com/kirel/detexify
and there is sample data to train from for #AI and #ML researchers:
🔗 https://github.com/kirel/detexify-data
@cappallo pretty much agree; but I remember going through list like these
https://oeis.org/wiki/List_of_LaTeX_mathematical_symbols
or
https://upload.wikimedia.org/wikipedia/commons/9/97/Comprehensive_LaTeX_Symbol_List.pdf
until a knew a few hundred of them by heart.
Detexify would probably have saved me a bunch of time if I knew it in my early LaTeX life.
what????????🤯
This is extremely useful, thank you!
@msokolov It does exist (and its linked from the Detexify FAQ btw).
Maybe i'll make a nice webm post like this one for this unicode tool :)
Attached: 1 image People seemed to love yesterday's #tool and asked for a #Unicode version it. Here it is! 🔗 https://shapecatcher.com It does not work as well imo, but it's good enough. Other issues are that it does not seem to be actively maintained since 2012 and it is not libre and open source. The dataset does not (understandably) cover the whole Unicode standard. #AI #ML
@laufi yup und geteilt. Wird in jedem meiner Workshops erwähnt, der Mathekrams macht.
Und ich hatte den Toot Autor extra gebeten TeXLaTeX hinzuzufügen, damit man es gut teilen kann :) Erinnerung da schadet ja nie.
@kirch I've posted about it earlier today: https://sigmoid.social/@solalnathan/109449977421049615
Don't hesitate to share it ;)
Attached: 1 image People seemed to love yesterday's #tool and asked for a #Unicode version it. Here it is! 🔗 https://shapecatcher.com It does not work as well imo, but it's good enough. Other issues are that it does not seem to be actively maintained since 2012 and it is not libre and open source. The dataset does not (understandably) cover the whole Unicode standard. #AI #ML
@Hearth i also made a post about it earlier today: https://sigmoid.social/@solalnathan/109449977421049615
Too fast 
Attached: 1 image People seemed to love yesterday's #tool and asked for a #Unicode version it. Here it is! 🔗 https://shapecatcher.com It does not work as well imo, but it's good enough. Other issues are that it does not seem to be actively maintained since 2012 and it is not libre and open source. The dataset does not (understandably) cover the whole Unicode standard. #AI #ML
It is very cool.
What about searching for how to make different kinds of stacked forms like matrices, or case statements?