Want to be astounded?

http://grail.cs.washington.edu/projects/videoenhancement/videoEnhancement.htm

What a waste of time.

I think it's pretty neat, actually.

Me too, very cool stuff.

That was pretty cool.

Yes, making video game engines is so much more productive, Leadwerks!

Object removal.

Wow. Very clever techniques.

I thought that was amazing.

Clever trick :)

Must be using some advanced pattern recog to be able to do that.

wow. interesing...I like the 'object removal' thing.

While we're talking about graphics papers, does anyone know if there's a site or blog collecting videos from this year's SIGGRAPH presentations? I've been on youtube, but it doesn't look like there's very much there at the moment.

Yeah the 'object removal' was cool. I wonder if they do 'object insertion'?

This actually is a very valuable asset for documentation of Artifacts and hard to reach places where conventional camera crews cant reach. I can see this technology making its way into the Tv/Media very soon (Budget Budget Budget)

Otus:

No, but Microsoft's tech does:
http://gizmodo.com/5036270/microsofts-unwrap-mosaics-add-mustaches-and-more-to-video

that is some amazing tech!

Very very cool

Thats pretty cool. Shame cameras mostly suck at taking acurate values, especially in shadows. The human eye sees a lot more detail in the darks and tones down the highlights more so we get a much broader range than normal cameras see. Plus cameras often get the colour temperature wrong, especially missing reflected light outdoors making them really dark and cool (bluish) where often outdoors they are much warmer (especially noticeable under things like arches).

Still pretty good tech though.

Well, mustache is a bit lame. I was thinking like a brand new Lamborghini on the parking lot. Maybe use that as a tennis net to practice serving.

Does anyone understand the "spacetime" magic he mentions there. I mean what are they talking about there? are they bending time and space to be able to stabilize the effects the are applying :D lol

swift offer me an explanation here of what you think they ment, I'm sure you have atleast a theory about it ;)

I think this image has soemthing to do with spacetime of existanse effect they talked about and also with Einstans theory of spacetime and gravity pulling effect of large objects...



cool stuff

Now this is what Einstain had in mind when he announced timespace continuum, right this thing here:

http://research.microsoft.com/unwrap/rkrf_short.wmv

All hail Eainstin ;)

Maximo:
What they are doing is they are computing the depth of each pixel in the video. This gives them a cloud of points which is kind of like a voxel representation of an object in 3D.

They then project the high resolution images onto that model, and re-render the video. They also do some funny stuff with how they blend the two sets of data together so as to smooth things out and retain the moving specular highlights.

(Also, interestingly, just the act of computing that cloud of points should give them the ability to up the resolution of the video to some degree.)

Spacetime Fusion is just some fancy name they've given this. Because it works in both space (the image space of a still frame) and time (interpolated in position over the course of the video).

And that's about as good a description as I can give you without fully understanding what they're doing. :-) I know a bit about cleaning up video with temporal filtering and how they calculate the point clouds, but there's a lot of stuff they're doing here which is new to me, and I don't know how they're doing it. Like the specular highlight stuff.

The mustache thing in comparison seems fairly straightforward. :-)

> All hail Eainstin

At least learn how to spell his name then.

dude I will spell it any way I want!

Einstin, Ainstain, Einstein, Einstino, Enstin, Etesting, or what ever way I preffer, you will still understand what I mean so please no lecturing here about spelling :P

Or any three ways, if you want to be technical.

Is this the right to 'free spelling' now?

Sounds absurd but reminds me of the recent call to introduce variant spelling in the UK for certain words that students commonly have problems with. Talks of adding common alternative spellings in the official dictionary.

Seems a bit silly to me. Next they will want to introduce texting lingo and weird case spellings.

Well maybe not free spelling then, just spelling enhancements.

European Union Spelling Enhancements

The European Commission have just announced an agreement whereby
English will be the official language of the EU, rather than German, which
was the other possibility. As part of the negotiations, Her Majesty's
government conceded that English spelling had some room for improvement and
has accepted a five year phase-in plan that would be known as "EuroEnglish".

In the first year, "s" will replace the soft "c". Sertainly, this will
make the sivil servants jump for joy. The hard "c" will be dropped in favour
of the "k". This should klear up konfusion and keyboards kan have 1 less
letter.

There will be growing publik enthusiasm in the sekond year, when the
troublesome "ph" will be replaced with the "f". This will make words like
"fotograf" 20% shorter.

In the third year, publik akseptanse of the new spelling kan be
expekted to reach the stage where more komplikated changes are possible.
Governments will enkorage the removal of double letters, which have always
ben a deterent to akurate speling.

Also, al wil agre that the horible mes of the silent "e"s in the language
is disgraseful, and they should go away. By the fourth year, peopl wil be
reseptiv to steps such as replasing "th" with "z" and "w" with "v".

During ze fifz year, ze unesesary "o" kan be dropd from vords
kontaining "ou" and similar changes vud of kors be aplid to ozer
kombinations of leters.

After zis fifz year, ve vil hav a realy sensibl riten styl.
Zer vil be no mor trubls or difikultis and evrivun vil find it
ezi to understand each ozer.

ZE DREAM VIL FINALI KUM TRU!

Something tells me that it takes their system a long time to generate all the info it needs to allow them to do what they want (object removal would require a buttload of information for it to 'guess' what's properly shown behind it in the case that the camera never rolls its view over at one point to actually show the contents behind it)

Still some very cool stuff, could save a crapload of time where one would either have to retake a shot or manually edit things out frame-for-frame to make something more suitable.

erm..go on then, astound me.

lol

Sure, why not? Then maybe next we can make 3,4 and 5 all valid answers to 2+2. I mean, it'll stop all those people who get it wrong from having to deal with the stigma of being wrong. I know it's been a few years since you left Britain, so it probably does seem silly, but it's the British way now. I'm pretty sure we'll be giving foreign language degrees to people who speak slowly and loudly in English pretty soon. That'll boost our flagging results in languages.

hmm, this is interesting.

I'm particularly interested in how they generated the stereoscopic/3D data- i've been toying with this recently and i've only been able to detect 3D by checking differences between the current and last frame- and the larger the "border" around the difference, the closer it is....

Sinu:
I asked if you wanted to be astounded, I didn't say I would do so. :-)

that's groovy cool! Couldn't they combine the depth proxies and filter them to the point where you could generate a textured 3D model of an object/scene using a 360 vid from a camera??

Very interesting video. One use of this could be to improve video compression. You could have high res keyframes and low res delta frames and retouch the detail in between. Nice.

That's very clever.

Wow! I so wish it was an out-of-the-box Premiere or Vegas plug-in :D

I'm going to assume that this is going to filter through to Hollywood very quickly indeed.

The camera stabilisation alone would float any company providing that service.

Some really amazing stuff.