How do I get into the AI field?

Attached: 1542171692033.png (555x464, 333K)

wouldn't they need cameras with a much better resolution to even capture the faces accurately enough to differentiate between the asian same face

Those camera's already existed for a long time grandpa

First, get a firm understanding of calculus. Then start reading about how a basic neural network works. Write your own implementation. Think about how it could be improved. Improve it. Then start reading some more advanced material.

op's pic looks like crap, which is what I was referring to
no sane person would drop tens of thousands(possibly more) for ultra-res CCTV

>neural networks
>calculus
Jesus, just stop it

>implying the chinese have AI
its just a room full of gooks with assburgers who are really good at remembering faces

Computer can easily detect that if they machine-learn it with gook faces. I can't even differentiate between wh*tes.

it's not FOSS you blanket retard, the raw high resolution capture data is proprietary

chinese facial recognition is not small feat

My Thinkpad can't handle it. It overheats and shuts down even on toy examples.

tensorflow is the easiest way to get into machine learning since python only takes a few months to get the hang of for a complete beginner.

This is the correct answer

>cs brainlet can't even handle calculus

Attached: 17800461_201105857050663_1727969891973630656_n.png (570x960, 661K)

apparently they're able to ID people with some stupidly high accuracy by analyzing the person's walking cadence/pattern

Every fucking scare story by the MSM always now features images of people with squares around them.
Every time.
The care story and its predictable graphics are getting tedious

this is why I cycle between walking, running, crawling, moonwalking, and cartwheeling across the street. catch me if you can, gooks

chinese propaganda. If it was that easy google would have already done it

>its another chink ai acam
yikes
ai will never amount to anything significant

Attached: 1560436398102.jpg (500x667, 411K)

It already has amounted to something significant, and there's no signs of it slowing down.

Tbh I don’t think calculus is necessary to get an intuitive understanding of what convolutions do. If one were to develop ML applications, blackbox understanding of the background math may be a that’s totally needed. But when seeking to contribute to the field of AI, absolutely, calculus, differential equations, and practice each with vectors rather than just scalars.

What makes you think they haven't?

Because google likes making money.

I smiled.

Can you prove that intelligence is really anything other than if/else statements?

>oh look there's the retard
pretty easy to catch you user

It is all based on derivatives and gradients, stupid faggot

Linear Algebra, Fast Fourier Transforms and Probability Theory.

I chuckled

Open the gate and walk through you fucking retard.

>all gook faces look the same so they have to analyse their walk
top kek

>Probability Theory

literally an astrology

>literally an astrology
Probability theory is just calculus with non-Lebesgue measures.

If it is astrology so is all of physics and engineering.