Recent Posts

Pages: 1 [2] 3 4 ... 10
General Chat / Oh baby! Just look at you now...
« Last post by Art on April 27, 2017, 11:00:28 AM »
Artificial Intelligence, Test Tube babies, Surrogates and now, the Artificial Womb!

This has mind blowing possibilities for the future.
General AI Discussion / Re: outline from gadient mask
« Last post by korrelan on April 27, 2017, 09:43:16 AM »
Basic Feature Detection

With any kind of feature detection you are basically looking for a pattern within a pattern.

The simplest method is to apply a line/ outline convolution filter to your image and then search for known recognisable pixel patterns. The simplest pixel pattern is a 3x3 block (fig A); there are various ways of defining the block but the simplest is to assign a binary value to each surrounding pixel from X/Y.  So 20 using this encoding method would equal a P1/ top left right angle. 

This is the method I used do get the outlines of the shapes in this vid. Except I use the recognised binary shapes to move a point around a perimeter. I could easily mark the corners of shapes by noting the binary numbers for corners. Etc.

A 5x5 (top right) or greater dimension block can be used but you would probably need a different encoding method; storing the pixel vector positions relative to the centre x,y as s simple list of vectors to be checked for example.

You then scan your image, summing the binary values of the surrounding pixels of each x,y point; the value returned can be easily checked against an array of known shapes (P1). Store the positions/ vectors in an array/ stack. This method is easy because you are just checking a one byte (0 to 255) value against a know array of byte/ shape values.

Shape Recognition

Once you have all the vectors for the found feature patterns and their relative positions to each other you can then find shapes (fig B) by checking for alignments etc.

Eigen Vectors

This is another method for checking one pattern against another without using an outline convolution filter; though usually a contrast filter is applied to bring the values of the image being checked in line with our set of eigen features

The RGB or Greyscale values of the surrounding pixels of the x,y point being scanned are subtracted from the pixel values being checked and run through a linear distance formula to see how similar they are. Rather than a simple binary array of numbers to check against; it uses small shaded bitmaps similar to (fig A top right).

‘s1=greyscale for position 1 (on binary grid above)
‘c1=greyscale for position 1 from the image being checked.

‘do same for all nine surrounding pixels storing V#

.dist=SQR((v1*v1)+(v2*v2)+(V3+… for all nine surrounding pixels.

The dist returned is a measure of how similar the block of pixels being checked is to the block of pixels in our stored array of eigen corner shapes.

This method has the advantage that ‘eyes/ mouths’ can be defined as small blocks of pixels that can be checked against an image.

Motion Detection/ Optical Flow

If you have the relative positions for known eigen features within an image you can check them against the next image in the video stream and measure the displacement to log what’s moved/ how far and how fast.

Other Methods

There a loads of other methods or finding corners/ features in images. You could for example detect pixel changes around a circumference from your x, y scan point (fig C) using the relative angles to find corners.

You could use the binary method to get a rough idea where the corners are and then apply a more precise method to each found vector to weed out false positives.

My Method

Because my AGI is based on the human connectome/ nervous system I use a model of the human visual system to detect features.

Neurons in the AGI’s visual cortex become trained to recognise lines/ corners etc through experience and only fire when their receptive fields detect their chosen pattern of inputs.  This is like running several convolution filters at once as scale/ rotation invariance/ gradients and movement can all be learned by the same V! cortex model.

General AI Discussion / Re: outline from gadient mask
« Last post by yotamarker on April 27, 2017, 04:43:58 AM »
nope its 2^81
General AI Discussion / Re: outline from gadient mask
« Last post by yotamarker on April 27, 2017, 04:37:44 AM »
I have a question...
an image of 9 by 9 pixel has 512 possibilities ?
Graphics and Video Software / Re: Thanks Again
« Last post by on April 27, 2017, 01:20:30 AM »
Apologies for making it feel an apology was necessary.

Feedback and critiques are welcome and encouraged.

That is the whole point of sharing fan created content.

Art said,  "you often seem to spread the moment of what is promised, somewhat thicker than that which is delivered"

You are right about that.  The reason for this is that some A.I. research is not easy to deliver.   PC Controlled talking robot heads that mechanically lipsync to text to speech, for example is a lot of fun to research, but it is not easy to deliver.   Maybe that explains why it is more commonly found in fancy research laboratories.

Try rigging a bone system into the female android SVG file above.   Doable definitely.  Not suggesting members here can't do it, but still this is an example of research that is tricky to figure out and deliver on, I think. 
Home Made Robots / Re: The ALTAIR Robot Chronicles
« Last post by Tony Ellis on April 26, 2017, 12:39:06 PM »
Art, It was strange when the film crew brought the Pepper to our studio the guy controlling it was moving it manually (by hand) to filming positions. Peppers top half moves very smoothly and is very impressive, although it took some pre-programming before it did anything. The ALTAIR was working mostly by itself, I had pre-trained Peppers image into the ALTAIR via photos so he recognised the Pepper robot and did a welcome routine. The film crew (and Pepper operator) were impressed with the ALTAIR functionality.

General Chat / Re: Member Channels
« Last post by Don Patrick on April 26, 2017, 11:53:54 AM »
Sure, I'm always happy to spam:
My blog. Analyses of Turing Tests, Loebner Prize and other A.I. challenges.
My Youtube channel. Personal programming and robotics projects.
General Chat / Member Channels
« Last post by Freddy on April 26, 2017, 06:08:24 AM »
Since a lot of you are like me and have accounts all over the web, I thought it would be a good idea to start a topic where we can add our links.

So here's me :

I have a few user names and I am trying to make them a little more organised. In gaming I tend to use the name Kaywulf these days and for my work I usually use Widdershins - these may cross-talk !

My YouTube Channel (avatar work and other stuff) : Widdershins YouTube
Blog (graphics work, Daz, Unity etc) : Widdershins Studio
My Google+ Page (auto updates from the blog) : Widdershins Google+
Facebook Page (auto updates from blog) : Widdershins Studio Facebook
Discord Chat Channel : Widdershins Chat Channel Invite

The Discord one I only just joined but am really enjoying the chat system they have. I decided to try it as I am following a couple of games that I want to keep up with. The link I posted is an invite, so if you want to get in touch with me I will try to hang out there. Once registered (simple) you can use it in a browser or download a client.

I do use Facebook too, but mostly privately and I haven't got time to work out their privacy stuff so won't be sharing it, but there's the studio page there that gets auto updated with stuff.

So you're spoilt for choice if you have an interest in my dabblings and musings  ::)

Would like to see what everyone else has going on :)
General AI Discussion / Re: outline from gadient mask
« Last post by yotamarker on April 26, 2017, 03:50:22 AM »
I hate opencv so much I can't use it on and it just appears in too many search results like some retard commercial. the same goes for matlab.
General AI Discussion / Re: outline from gadient mask
« Last post by yotamarker on April 26, 2017, 02:09:05 AM »
that's the problem I don't understand it the way wiki explains it
Pages: 1 [2] 3 4 ... 10


Please login or register.

Login with username, password and session length
Meet the People Who Train the Robots (to Do Their Own Jobs)
by Tyler (AI News )
Today at 10:48:18 AM
Russian Military Robots
by Art (Robotics News)
April 25, 2017, 06:14:30 PM
by Art (AI News )
April 19, 2017, 06:19:36 PM
A Murder Posted on Facebook Prompts Outrage and Questions Over Responsibility
by Art (AI News )
April 18, 2017, 10:12:52 AM
AI based 3d scanner
by Art (AI News )
April 18, 2017, 10:04:39 AM
Artificial Earthling
by Art (AI News )
April 14, 2017, 03:49:08 PM
Faster websites with fewer bugs
by Tyler (Robotics News)
April 13, 2017, 10:50:45 PM
New Tools Needed to Track Technology’s Impact on Jobs, Panel Says
by Tyler (AI News )
April 13, 2017, 10:50:43 PM

Users Online

21 Guests, 2 Users
Users active in past 15 minutes:
Freddy, kei10
[Trusty Member]

Most Online Today: 31. Most Online Ever: 208 (August 27, 2008, 08:24:30 AM)