Can someone explain this image?
Collapse
X
-
UMMMM... I can't see anything but her boobies going in a curious direction, sorry LOLComment
-
Let me show off cognitive image recognition that will be built into Node.XXX image search.
{
"url":"http://i.imgur.com/7ZE5RCK.jpg",
"results":[
{ "score":"38.17%", "text":"old man sitting" },
{ "score":"12.08%", "text":"a sandy beach" },
{ "score":"37.10%", "text":"nude female running" },
{ "score":"6.32%", "text":"daytime" },
{ "score":"6.32%", "text":"monochrome" },
{"runtime":"1.0293s"}
],
"description":"a black and white image of an old man is sitting on a sandy beach and a nude female is running in the day time"
}Comment
-
-
-
Just another day in paradise.Success stories of porn webmasters - one click away:
Discover how top porn sites grow their organic traffic!
<span style="color:White">[email protected]</span> | Telegram: https://t.me/LT7_DigitalComment
-
Node.XXX just has text search at the moment - https://node.xxx
I am working out a few bugs on Image Search and expect it to be ready before the end of the year. So far results are really good but not quite ready to be rolled out in production yet. Although I may have a beta available on the site by October.
Interesting facts: Node.XXX has around 100 million web pages crawled and growing. We restarted indexing last week so the index only has 42 million web pages. However the Image Crawler has crawled a staggering 235 million images and I expect it will be half a billion by early November. Indexing the images will take two stages, the first is storing the images in a searchable database (index), the second is applying cognitive recognition to enhance image search using text.
Unfortunately we don't have the expertise or budget to do reverse image search yet, but we're looking at some third party options for integrating that next year.Comment
-
natural boobs are better ?Beautiful And Usable Web Design Creations For The Adult Industry Since 2003
I'm Yuu, Designer and Content Producer
Paysites - Affiliate Programs - Dating & Cam Sites - Mainstream Projects - Tube Sites - Banners - Wordpress Themes - NATs integration - Landing Pages
Check my Portfolio and Content Production OffersComment
-
We're not using a third party API for our image recognition. Neither Google APIs or Bing APIs are allowed for use with competing products (especially those using third party advertising) and they would be much more expensive than doing it ourselves.
Bing & Google both do much better image recognition than we can, but they have a budget a million times greater than ours.Comment
-
-
Perhaps it's a thought bubble he is having as to who he really feels he is inside as he sits in thoughtful prayer 1/2 before he is required to be at the hospital for the scheduled Caitlyn Jenner Special surgeryComment
-
Comment
-
http://lens.blogs.nytimes.com/2013/0...rom-obscurity/
"He stabbed and killed the man, then flung himself off the balcony. Ms. Aldag found him in the snow."
says KGB, but seems weird as she was able to leaveComment
-
-
Comment
-
Big tits swing from side to side in unison when running.
https://www.youtube.com/watch?v=iJve6j1ZhMo
Comment
-
PRACTICAL PHYSICS by BLACKMONSTERS
Newton: An object in motion tends to remain in motion along a straight line unless acted upon by an outside force.
blackmonsters: Big tits swing from side to side in unison when running.
I know which one I'm voting for.Comment
-
He couldn't catch the girl so he positioned himself ahead of her. Shortly the cane will slide out and trip her and then hardcore midget anal invasion will proceed.
Comment
-
bwack monstah power
Comment
-
She is Hot.
Our Services: Website Design | PSD to Responsive HTML5 | Logo Design | PHP/Custom Development | WordPress | E-commerce Development
Website: https://adultdc.com | Contact on: [email protected] | ICQ: 676735786 | Follow us on TwitterComment
-
-
-
It's a girl running on a beach going past a small person. Nothing else to explain.Comment
-
It's is only to show the contrast between the hope and energy of youth against the implacability of old age and deathComment
-
where did the image description came from ?
how are you doing it ?
is there api ? for this ?
how much for api access for something like that ?
yours description is better than google's visionComment
-
where have I seen this before?? oh yeahBig tits swing from side to side in unison when running.
https://www.youtube.com/watch?v=iJve6j1ZhMo
Comment
-
The cognitive image recognition system we are putting together for Node.XXX
The specifics are secret sauce. The specifics are also changing as I further my understanding of cognitive systems and keep developing.how are you doing it ?
Generally we examine the image and similar images in our database and feed into our system data such as alt tags, page content, image content of similar already described images then that's all compared with objects we already are able to detect with levels of certainty returned in the JSON results.
Yes but it's private.is there api ? for this ?
We haven't thought about providing this to the wider world.how much for api access for something like that ?
So I noticed, good isn't it ?yours description is better than google's visionComment
-
thanks for your reply.The cognitive image recognition system we are putting together for Node.XXX
The specifics are secret sauce. The specifics are also changing as I further my understanding of cognitive systems and keep developing.
Generally we examine the image and similar images in our database and feed into our system data such as alt tags, page content, image content of similar already described images then that's all compared with objects we already are able to detect with levels of certainty returned in the JSON results.
Yes but it's private.
We haven't thought about providing this to the wider world.
So I noticed, good isn't it ?
one thing i still dont understand,
description came from the source site where the image has been used by their alt or title tags
or
your code/program can understand the image and process it only by scanning image.
?
thanksComment
-
No the system analyses an image and uses many different data points to understand what the major things in the image are, scores them, then produces a result with scoring.
So the only input required is the image data then the system will use it's processing capability to try to understand what is going on in the image and what feature objects are present in the image.
However you will note from our scoring that "monochrome" has a very low score, yet it's obvious the image is not colour. The reason for that low scoring is that we need to compensate for those kinds of results, otherwise "monochrome" would be the major aspect and returned percentage of the image and skew the result.
Google use machine learning to do this, obviously they have huge resources at their disposal, so I am using novel techniques to try to do something similar drawing on a number of already written open source projects and putting them together to create a much less expensive (in dollar terms and processing expense) system that can produce fairly good results.
I have been playing with this kind of stuff for many years. With varying degrees of success.
This is an ongoing project, but image search is very high on my priorities for Node.XXX. People are using image search more and more these days and nowhere can I think of it being more important to relevant search results than in porn.Comment
-
that is some awesome stuff.No the system analyses an image and uses many different data points to understand what the major things in the image are, scores them, then produces a result with scoring.
So the only input required is the image data then the system will use it's processing capability to try to understand what is going on in the image and what feature objects are present in the image.
However you will note from our scoring that "monochrome" has a very low score, yet it's obvious the image is not colour. The reason for that low scoring is that we need to compensate for those kinds of results, otherwise "monochrome" would be the major aspect and returned percentage of the image and skew the result.
Google use machine learning to do this, obviously they have huge resources at their disposal, so I am using novel techniques to try to do something similar drawing on a number of already written open source projects and putting them together to create a much less expensive (in dollar terms and processing expense) system that can produce fairly good results.
I have been playing with this kind of stuff for many years. With varying degrees of success.
This is an ongoing project, but image search is very high on my priorities for Node.XXX. People are using image search more and more these days and nowhere can I think of it being more important to relevant search results than in porn.
i mean computer can guess/pin point whats in the picture and where it is .
thanks for details.Comment
-
Comment








Comment