I need a way to gather all of the hoes on instagram. And sort

>>59262327

get a fucking life you clown

Try going outside instead and try talking to a girl.

>>59262354
>>59262360
Why are you losers assuming I'm doing this to meet women? Virgins.

>>59262327
You can measure the amount of skin exposed in each picture

>>59262327
Uh, this is too difficult for you to program.

Figure out how you can do it communally with sharing perhaps. Or just watch porn.

Fuck off, niggerposter.

>>59262327
> programmatically
>>> /b/

>>59262409

>Uncontrollable primal urges staring at a woman like pic related

Dudes like you end up fucking black trans hookers in the back of Volkswagens before shooting up a synagogue in a drunken stupor.

File: 1384209_592729767452323_2143132045_n.png (62KB, 245x245px) Image search: [Google]

62KB, 245x245px

>>59262476
>implying I'd ever get in a filthy VW

File: 1455714568174.gif (304KB, 500x375px) Image search: [Google]

304KB, 500x375px

>>59262327
>programmatically
cringe

buy a blow up doll and call it a day anon

Google Caffe and imagenet.

>>59262327
I'll give you a serious answer

Assuming you wanted this to be robust, in addition to heavy text analysis (beyond matching keywords, but looking at context) you would need a conv neural net that would be able to read an image for promiscuity (and again) tie that in to text context.

100% too much for you to do since you're asking this kind of question, but possible with effort.

>when niggers try to into technology

>>59262327
Use the percentage returned by isitporn.com
higher percentage => more lewd image

>>59262327
>>59262601
>inb4 OP "so what language should I start with"
>"I already know visual basic and html"

>>59262553
It's exactly how those net nanny programs do it

They search for sex-related images and text

Shouldn't be too hard desu senpai

>>59262376
We don't buy your lies, jackass

>>59262784
Those 100% do not leverage an intelligent neural net

Sure if OP was to make a program that determined 'slutiness' for humor alone, performing basic text analysis

But to make a good one..

>>59262553
I think there might be public nsfw training sets already. Could probably get really far just doing that and parsing their bio for a/s/l.

Getting a list of all users(or a lot) might be hard. Maybe go through each commentator and see who else they talk to.

>>59262327
Make a script that scrapes all their posts, apply one of those porn classifier NNs to each image, then sort them by who had the highest average probability for porn