cell phone pano

Discussion forum for images in Max Lyons' and other image galleries...post links to your own images here

Moderator: spamkiller

Post Reply
dsp
Posts: 586
Joined: Tue Feb 20, 2007 11:09 am
Location: Hoboken, NJ

cell phone pano

Post by dsp » Tue Apr 02, 2013 1:30 pm

Certainly a technically poor panel, but in the spirit of board activity, I'll post it. Quick 5 shot pano of the East Room of the White House a few minutes before Obama came in. He announced a Grand Challenge to understand the brain. The initial part of this will be a lot of technology development, with imaging playing a big role. Eventually, some of the tech will make its way into consumer cameras, and allow me to take nicer panos with my cell phone :)
cheers, dsp
Image

Jim Zuccaro
Posts: 71
Joined: Wed Apr 06, 2011 9:44 pm

Re: cell phone pano

Post by Jim Zuccaro » Sun May 19, 2013 11:48 pm

dsp wrote:Certainly a technically poor panel, but in the spirit of board activity, I'll post it. Quick 5 shot pano of the East Room of the White House a few minutes before Obama came in. He announced a Grand Challenge to understand the brain. The initial part of this will be a lot of technology development, with imaging playing a big role. Eventually, some of the tech will make its way into consumer cameras, and allow me to take nicer panos with my cell phone :)
cheers, dsp
Image

Darcy, in the old days, it was said that thermodynamics owes more to the steam engine than the steam engine owes to thermodynamics. I'm sorry for the 'dig', But I bet that the understanding of vision will owe everything to the (consumer) camera and computer business, and very little will be owed in the reverse direction, Challange included. I say that with the best wishes for you!

thanks, Jim Z

dsp
Posts: 586
Joined: Tue Feb 20, 2007 11:09 am
Location: Hoboken, NJ

Re: cell phone pano

Post by dsp » Thu May 23, 2013 10:24 pm

Hi Jim!
I'd say your dig is (mostly) accurate. From the tech side I see a few things coming down the pipeline that are initially research driven (as big research projects can afford expensive, low volume runs) that haven't yet found consumer markets, but at the same time, I see the push for megapixels and fps in the consumer market really enabling many things in research community. Right now, the machine vision community straddles the gap nicely.

We'll see how this whole thing plays out. Sequestration has taken a lot more out of science than this would put in...let's hope that the gov't can get their act together and come to some compromise that moves us forward!

Jim Zuccaro
Posts: 71
Joined: Wed Apr 06, 2011 9:44 pm

Re: cell phone pano

Post by Jim Zuccaro » Fri May 31, 2013 10:39 pm

Darcy,

I agree, the loss of science money from sequestration isn't good.

What I was thinking of is the 'computational theory of the mind'. When we try to program computers to emulate what our minds do, and reduce 'thoughtful' operations down to a series of smaller and smaller operations, ultimately reducing the 'intellection' down to a very many tasks small enough for thoughtless computer demons to do each of them, we understand how the brain may work the same way. Contemporary image processing seems to be following the same course. Developments of de-mosaic-ing, parallel processing, compression and sharpening have 'discovered' some of how our vision works. Pattern recognition algorithms, too.

It's the difference between the outside-looking-in examination of the mind, hydraulic analogies and pecking-pigeons behaviorism at the worst and inscrutable magic-boxes in the rest, vs the building-block-up emulation, creating or copying the functionality of the mind to do something necessary, difficult and useful (profitable).

thanks! Jim Z

PS: Any politician, of either persuasion, announcing a 'Brain Initiative' reminds me too much of the Scarecrow from Oz...

dsp
Posts: 586
Joined: Tue Feb 20, 2007 11:09 am
Location: Hoboken, NJ

Re: cell phone pano

Post by dsp » Sat Jun 08, 2013 11:34 pm

Hey Jim, you are right (or at least I believe the same thing!). One particularly difficult problem in computer vision, and as a result, neuroscience, has been in feature recognition that is consistent across many view or sections. I can't help but think that once we get an algorithm(s) that gets closer to the human response based "ground-truth", we have gained some insight to the probable underlying computations of the brain. This seems to be more easily connected to visual processing than some other brain areas, but nonetheless hold water across the board. Simply understanding what data the brain considers important already tells us a lot...
cheers, dsp

Post Reply

Who is online

Users browsing this forum: Baidu [Spider] and 2 guests