Slashdot Mirror
Algorithm Seamlessly Patches Holes In Images
Beetle B. writes in with research from Carnegie Mellon demonstrating a new way to replace arbitrarily shaped blank areas in an image with portions of images from a huge catalog in a totally seamless manner. From the abstract: "In this paper we present a new image completion algorithm powered by a huge database of photographs gathered from the Web. The algorithm patches up holes in images by finding similar image regions in the database that are not only seamless but also semantically valid. Our chief insight is that while the space of images is effectively infinite, the space of semantically differentiable scenes is actually not that large. For many image completion tasks we are able to find similar scenes which contain image fragments that will convincingly complete the image. Our algorithm is entirely data-driven, requiring no annotations or labelling by the user."
Uncensored Japanese pornography!
This message printed on 100% post-consumer recycled electrons.
Broken or flaky video files. Nothing is more irritating than an mpeg, etc error that causes an entire block to go black and smear itself all over the place until the next keyframe. I don't expect realtime correction, but it would be nice if I could patch the file rather than do another six hour encode.
May the Maths Be with you!
It was as if a million fake celebrity pr0n websites cried and were suddenly silenced...
My blog
And if you dont have any pictures database, there's always GREYCstoration:t ion/index.htmlt ion/demonstration.html
http://www.greyc.ensicaen.fr/~dtschump/greycstora
It's pretty impressive:
http://www.greyc.ensicaen.fr/~dtschump/greycstora
and works with the gimp.
I take a picture of a hole?
I prefer Flambe as apposed flamebait.
You never know what that "kinda-like" picture used to patch contains. You might get the opposite of what you want.
We used to have a Bill of Rights. Now, with the rights gone, all we have left is the bill.
CSI Miami and NY have had infinite zoom capability with photos for years, and you excited about this? Bah.
If brevity is the soul of wit, then how does one explain Twitter?
If any hole in the image can be filled with a part of another pic, can't you compress an image by replacing one piece at a time with a reference to a patch?
That only works if your patch addressing space takes less space than the bits you're replacing - and of course when you reload the image, you'll still get say a cat instead of an iguana in that window...
Also, how about replacing with patches of higher resolution than the original? I realize it would all be technically lossy as hell, but the compression artifacts should not be very noticable to the human eye, right?
I'm not sure you really understand the concepts here. Replacing a patch with a higher res would be possible (but you'd have to resample the image first, basically) - and would either be incredibly lossy or perfectly unlossy, depending on your viewpoint.
From a compression standpoint there's no reason to consider a high res replace as more lossy as anything else. From a recognition standpoint, whether you're doing it high res or not, this would be a method that throws out image details for others... but that doesn't have anything to do with the resolution. So this is a lossy image manipulation, but not really a compression...
And of course, none of that would cause any compression artifacts, so yeah the human eye wouldn't notice (assuming this software works as claimed)
So to go back over the concepts:
Lossy - a compression or manipulation to an image or other digital file from which you cannot reconstruct the original bits perfectly
Compression Artifact - a noticeable image tearing or other visual defect allowing one to differentiate between a lossy-compressed file and it's original
Additionally, how about using this for movie compression? Filling in based on info from previous and next frame.
That's how movie compression came about. The first moving-file format that was widely available was animated GIF - which quickly got onto the trick of using transparent pixels for non-changing parts of a scene.
MPEG (1) one upped it; one part of the spec specifies which blocks are to be sent in each frame; you can leave out any blocks you don't want... (they also smartly seperated the chrominance and luminance channels, and subsampled the chrominance channel - not only is it a smart compression as the human eye perceives luminosity at greater fidelity than chorminance, but it also ups the chances that you don't have to transmit some blocks)
Fast Forward to MPEG-4 (non-H.263) - same basic block structure, same ability to not draw blocks, and now you can even specify offsets for blocks - you have probably heard this technology referred to as motion compression - basically if something is moving on the screen but remains relatively the same pixel values regardless of motion, the movie file will record the motion without recording every pixel - the difference between a good MP4 compressor and a bad compressor mostly has to do with how well it identifies candidates for motion compression, is my understanding...
Instead of "You appear to be writing a letter. Should I format it for you?" I guess we'll get "You appear to be viewing Japanese pornography. Should I de-pixelate it for you?"
SJW: Someone who has run out of real oppression, and has to fake it.
This is very cool, and I wonder how similar it is to what the brain does with respect to blind spots?
For those who don't know: each eye has a surprisingly large blind spot at the place where the optic nerve enters the eye. At reading distance, in the right eye, it's about four or five inches to the right of the spot at which you are gaving, and many textbooks and "fun with optical illusions"-type books will have a diagram... like the one on this web page... and directions for finding it. The blind spot is much larger than the dot on that web page, incidentally. If you explore, you'll find that... at the distance at which the dot disappears... the blind spot is nearly an inch wide and an inch-and-a-half high.
Even allowing for the fact that each eye has the blind spot in a different place so they fill in for each other, once you discover how big the blind spot is... and how relatively close to your position of gaze it is... you'll be astonished that almost nobody notices it until it is pointed out.
The brain does something more or less like filling in the blind spot. I say "more or less like" because it is very hard to answer the question "what do you see in the blind spot." For example, if you hold a computer keyboard at the right distance so that you're looking at the "G" key and the "K" key is in your blind spot, what do you see? Certainly not a black spot, certainly not a white spot, certainly not a "hole" or emptiness. Probably you have an impression of computer keys. Do you see a letter K? Certainly not, yet somehow you don't see a blank key, either.
Incidentally, I used to suffer from migraine headaches, and one of the symptoms for some people is the formation of blind spots which can be even larger than the "normal" blind spot, and can appear in central vision. One one memorable occasion, I was looking at the cover of a hardbound book, and I can tell you that when I looked at the title, my perception was the stamped, printed title disappeared, yet I would have sworn in a court of law that I still saw the cloth texture extending across the blind spot.
Although he does not specifically refer to it as a migraine illusion, I believe Lewis Carroll was known to be a migraineur, and in Chapter V of Through the Looking-Glass, "Wool and Water," Alice notices that "The shop seemed to be full of all manner of curious things -- but the oddest part of it all was that, whenever she looked hard at any shelf, to make out exactly what it had on it, that particular shelf was always quite, empty, though the others round it were crowded as full as they could hold." Any migraineur who experiences central blind spots will recognize this description.
Hays and Efros' system--relatively-simple algorithm operating on a large database of previously-seen images--seems to me to be the sorta-kinda way in which one could imagine the brain working.
I wonder if there's any way to test this?
"How to Do Nothing," kids activities, back in print!