/sci/ - Science & Math

Computer vision is broad - I can recommend some books since I work in the field, but maybe you have something more specific in mind.
One popular mathematically minded text is "Multiple View Geometry in Computer Vision" by Hartley and Zisserman.
There's also standard CS degree texts on image processing (another understanding of what you can learn about images)
Deep Learning by Ian Goodfellow at all is a popular tome in ML, also used for CV purposes.
There's a short book by Nelson (physicist) which implements some basic python and I at one point toyed around implementing a python network that recognizes the NIST database digit images
see
https://youtu.be/QrJ9LyA-Z-U
https://youtu.be/z2aq21lMw40

>>12279707
>>12279709
>>12279731
The most interesting one for me is 2., but you actually missed to point out variants of it, which deserves another bullet point.
Namely the quote which is not an actual quote but a satirical transscription which is still first person.

E.g. someone says "to lose weight, eat at 400 cals deficit every day" and you reply with
>slowly starve yourself to death
As if that was the advice given.

>>12153059
Yes, I mean I posted it 20 posts above.
Although I won't do so if people are annoying about it, as in writing all cap posts.

>>12153047
Okay, although I still don't like the level.
Also do you want to do a project with people on github or just toy around on your own?

I really liked this paper, have a look if something is for you.
http://www.cs.ru.nl/~spitters/editorial.pdf

Or you like numerical analysis?

Do you see how to computably translate topology into coding? It's general tough.
The Heyting algebra stuff is highly relate to topology via frames (as in frames and locales), if you're into that
https://en.wikipedia.org/wiki/Pointless_topology

There's Lie theory libraries for C++, one could do something similar. I'm personally very interested in group interpolation, geodesic or otherwise, of two elements or more. There's papers on that.

imagine
[math] 0\to \mathbb Z\stackrel{\imath}{\to} \mathbb R \stackrel{\pi}{\to}\mathbb R/\mathbb Z\to 0 [/math]
was a splitting extension.
Let's make some theory where that's true. We can fuck with R if you like to achieve it.

>>11633892

>>9174806
>>9174833
but honestly this type scenario where good old america invented in the garage type story would be the absolute greatest story/history for future generations. similar to PCs only exponentially more important