Want to read Slashdot from your mobile device? Point it at m.slashdot.org and keep reading!

 



Forgot your password?
typodupeerror
×
AI Robotics Technology

Robot Produces Paintings With That 'Imperfect' Human Look 74

kkleiner writes "An artistic robotic system named e-David has been developed that produces paintings that appear to be created by humans. Using an iterative process of brush strokes and image comparison, e-David's assembly line welder arm can paint in up to 24 colors and add shading where needed. The robot even cleans its five brushes along the way, according to University of Konstanz researchers who developed the system as an exercise in machine learning."
This discussion has been archived. No new comments can be posted.

Robot Produces Paintings With That 'Imperfect' Human Look

Comments Filter:
  • Interesting (Score:5, Funny)

    by Anonymous Coward on Sunday July 28, 2013 @04:02PM (#44408237)

    how it models imperfection so perfectly

  • by Anonymous Coward

    http://en.wikipedia.org/wiki/Light_Verse_(short_story)

  • by arth1 ( 260657 ) on Sunday July 28, 2013 @04:14PM (#44408275) Homepage Journal

    This lacks one vital component: Creativity
    A painter may think "I may want to make that woman's eyes a bit more smiling", and then do so. Or think "If I add a stone fence between the buildings, it will look more severe".
    Or even "the sky would look better with a green streak".

    So while this might be a nice exercise in machine learning, don't insult its good workmanship by calling it art.

    • by Cyberax ( 705495 )
      A lot of "creativity" is overrated. Just wait before this robot can replicated that "just a bit more smile"...

      And what I'd like the best is the ability to reproduce paintings. I really like art, but I can't stand the 'artsy' types that claim that the original paintings are somehow magical. Sure, photographic reproductions are total shit but if this robot can be taught to make stroke-for-stroke reproductions - it'd be priceless.
      • A lot of "creativity" is overrated.

        QFT. It's not the same with paintings, but most people find modern (popular) music creative. AFAIK it's not done by robots yet, but most of it is taking samples from other songs, writing lyrics according to known formulas, and autotuning the singer's voice.

      • Let me know when the robot starts deciding what to paint, and then we'll talk about overrated "creativity" is....
      • Copying stroke for stroke is a different thing altogether. There is a whole industry for this. http://www.artsstudio.com/ [artsstudio.com] Price ranges with quality. Genuine paintings done by hand go from $200 to somewhere around $10,000 to $15,000 I think. They are not priceless. There is something about human nature the values the original. The price of art is a pure economic ideal. It is worth exactly what someone is willing to pay for it, so you can't really argue that someone overpaid.

        The high end copies entail using t

        • by Cyberax ( 705495 )
          Yes, I have several reproductions (clearly marked as such) of well-known pictures and I refuse to buy original art except for works of no-name artists. Good reproductions are expensive and still are not perfect. Besides, most artists prefer to paint not the exact replicas but something "in style of" them.

          I believe, that creating a system that can analyze the precise colors (using a spectroscope) and a robot that can mix pigments to produce the desired reflection spectrum is absolutely feasible. It'll be e
      • by The Cat ( 19816 ) *

        A robot will never create art. They are machines. Art requires a soul.

        • And machines can't have souls because... ?

    • by asmkm22 ( 1902712 ) on Sunday July 28, 2013 @04:28PM (#44408339)

      You should read the article...

      It specifically mentions how the machine is not remotely "creative" or even trying to mimic creativity.

    • by hazem ( 472289 ) on Sunday July 28, 2013 @04:30PM (#44408345) Journal

      When dealing with most visual art, you're restricted to viewing the end product. If I go to the Louvre or the MOMA, I can look at the finished products but cannot see the process by which they were created. These paintings, for the most part, are "art", based solely on their end-state; and the fact that they are in a museum of art.

      So what happens when you have a painting made by a machine put up in a gallery next to a painting done by a human being, and you can't tell which is which? A "Turing Test" of sorts. What if you hook the viewers up to an FMRI and see that both paintings generate an equivalent emotional response in the viewer?

      If the machine-made painting is "not art" because it was made by a machine, what does that mean for human-made painting? Is it no longer art because it was indistinguishable from something that we've determined is non-art?

      At that point, what is the definition of "art"? And the criteria for determining what is and is not art?

      Do you remember that guy who had paint forced up his rectum as an enema, and then he stood over a canvas as is sprayed back out? This was considered art (by the artistic community). If that meets the standard for "art" then I'm willing to give a robot (and its creators and programmers) the benefit of the doubt.

      • And the criteria for determining what is and is not art?

        The key criterion is creativity and expressive/emotional content, and wherever that comes from, that's the source of the art.

        In this case, some of the robot's paintings are quite artistic... but it's not the robot that selected the subject, captured the right feeling, chose a composition that accented it, etc. What happened here is that the robot reproduced some artistic images that were created by a human. Art? Sure. Robot art? Nope.

        In the case of your hypothetical art Turing test, where only the end p

        • by Kjella ( 173770 )

          I swear that if you took random sunsets from Google Maps and turned them into artistic-looking drawings/paintings they'd pass the "Turing test" with flying colors without any human being directly involved in the capture or composition. With all the bizarre things called art, it's almost impossible to say something was not somebody's "creative vision" even if it's actually a random machine-picked choice.

          The key criterion is creativity and expressive/emotional content, and wherever that comes from, that's the source of the art.

          In my experience it seems to be far more the audience's ability to project their expressive / emotional co

          • by shia84 ( 1985626 )

            I swear that if you took random sunsets from Google Maps and turned them into artistic-looking drawings/paintings they'd pass the "Turing test" with flying colors without any human being directly involved in the capture or composition.

            Who tells the machine to take a sunset? Who enables it to choose? The artist.
            This robot and any software picking & repainting google images is exactly as intelligent as the painters brush, just a bit more complex, and has no more self-initiative or creativity than a piece of wood.
            Taking a picture with an expensive DSLR doesn't make the camera the artist, and mounting it on a self-driving car that randomly takes snapshots still gives all the credit to the person that built this.
            And photographing the Mona

      • I've always wanted to see something like this happen. Get a robot to make a bunch of paintings, but tell everyone they were painted by a person. Art world goes crazy saying how great and original they are, then release the news that they are really just done by computers, and ask the art world to explain their initial reaction. That, or have something like the next big popstar just be a computer generated model. Have other people write the songs (they do anyway), and have the voice provided by someone wit
      • by stymy ( 1223496 )
        The paintings made by a robot may indeed be art, but the artist is the person (or people) who wrote its algorithms and gave it the seed data (a photograph or whatever) that resulted in the painting. Until a robot can come up with good composition, the real artist is still the one who took the picture it used.
        • by hazem ( 472289 )

          One of the coolest examples of machine learning was the TD-Gammon program done by Tesauro at IBM (http://en.wikipedia.org/wiki/TD-Gammon). What makes it remarkable is that while Tesauro wasn't a particularly good Backgammon player, he managed to develop this program that "learned" how to be a great Backgammon player - in fact, it learned strategies that no human players had ever tried before, "TD-Gammon's exclusive training through self-play (rather than tutelage) enabled it to explore strategies that huma

    • So if you were to view a selection of e-David drawn pictures mixed in images drawn by human artists, with no prior knowledge,how would you differentiate them? Isn't art in the eye of the beholder?

    • Most of what is referred to as "art" is merely the random scribblings of pretentious, cultist douche-bags. I'll accept the lack of creativity if e-David continues to lack those traits.
    • don't insult its good workmanship by calling it art.

      Nobody did.

  • via the telegraph wire was invented over 150 years ago. It predates voice telephone calls. This machine adds nothing new with its plotter which was invented invented 50 years ago and was immediately hooked up to both wireless and wired transmission at the time.

    BTW, the shadow of that tree is physically impossible (and no that doesn't make it "art".)

    This "invention" is total phail.

    • by Sique ( 173459 )
      First, it's not an invention. Second, it does not strive for a perfect reproduction (otherwise we could just take a photo). Third, no one claims this is art.

      So your rant: total phail.

    • I've got to say, I'm surprised at the negative comments thus far. A lot of engineering creativity obviously went into creating this robot, do you people not appreciate that? Do you have any idea how hard it is to get something this elaborate to work reliably? Have any of you naysayers actually tried to build even a simple robot?

      You call it a simple plotter. Really? A plotter is basically a dumb printer. This is is not a dumb printer. It creates the image iteratively by examining what it looks like and mo

      • by pigiron ( 104729 )

        "BTW, the tree's shadow would be roughly consistent with the sun being about 15 degrees above the left horizon."

        You are in serious need of a guide dog.

    • It predates voice telephone calls.

      Actually, it's NSA that predates voice telephone calls. Oh, wait, you meant the other kind of predating.

  • by bill_mcgonigle ( 4333 ) * on Sunday July 28, 2013 @04:27PM (#44408331) Homepage Journal

    I don't mean to take away form the robotics work or the research, but the headline appears to be jumping the gun. Most of the sample paintings look like GIMP filters or that machine at Chuck-E-Cheese that draws the kids' pictures while they wait.

    I was expecting a flexible arm mimicking Monet's technique or something. At this point I'd be much happier with an elephant painting [youtube.com] on my wall - it's more "human" than the robot's.

  • This is the sort of seemingly trivial machine learning achievement that will ultimately coalesce with other seemingly trivial achievements in the field to serve as the bulk of a future, human-level or above AI. Or so I think.

    The question is: will we know it when we see it? If we ever do develop a truly sentient AI, will we even be able to prove it?
  • by jjp9999 ( 2180664 ) on Sunday July 28, 2013 @04:44PM (#44408399) Homepage

    Interesting, but it's still not that much different from a printer with an algorithm to imitate a painterly look. There is software (like Corel Painter) that can transform photos to look like they were painted using different mediums. I could load a photo, use an automated feature in Painter, and print it, and it would basically do the same thing as this robot.

    • One could code a virtual Rubik's cube solver and do it all on-screen with a virtual cube. But isn't the hardware implementation so much more cool? http://www.youtube.com/watch?v=_d0LfkIut2M [youtube.com] So yes, you could pass your image through [FAVORITE PAINT PROGRAM] but that's boring. What's interesting about this robot is the implementation, the fact that you could have produced a similar final image at home by clicking a button isn't relevant.
  • by lxs ( 131946 ) on Sunday July 28, 2013 @04:45PM (#44408403)

    Robot has a hobby
    Builds Volkswagens by daylight
    Paints people at night

  • by Anonymous Coward

    It doesn't look any more natural than a host of 20 year old Photoshop effects.

  • by Blaskowicz ( 634489 ) on Sunday July 28, 2013 @04:57PM (#44408441)

    That robot will be rejected from a Fine Arts academy, failing the entrance exam twice, and from disgust and despise will try to take over the world instead.

  • I actually went out and read/watched TFA (gasp!) What is created by this robot is aesthetically pleasing to me. My definition of art is: something that can invoke an emotional response. Some of the pictures invokes loneliness, or struggle. I would call it art.
  • by pubwvj ( 1045960 ) on Sunday July 28, 2013 @05:21PM (#44408551)

    The irony is that while the robot has been perfecting the human look my wife has perfected that machine look. Her drawings look so real that people mistake them for photographs. She does edit out and add in but it comes out looking so real it is mistaken for reality.

  • I will not consider robots human-like until they eat paste.

  • Sorry to see no mention in TFA regarding Harold Cohen's "Aaron" drawing program which has been doing this for decades: http://www.scinetphotos.com/aaron.html [scinetphotos.com] This appears to similar albeit with the addition of a plotter + penholder. The debate of "is it really art or not" gets replayed yet again ...

  • As a someone with a Masters of Fine Art in painting, I can tell you there is not a lot of interest relating to art.

    First: "Our hypothesis is that painting ... can be seen as optimization processes in which color is manually distributed on a canvas until the painter is able to recognize the content" is off base
    All the lines in all the work are all the same length and thickness. Almost no artist simple distributes color. Artist chose details and focus.In this case David is being helped because it is using com

  • DRAW X+RND(10),Y+RND(10)

    Done.

As you will see, I told them, in no uncertain terms, to see Figure one. -- Dave "First Strike" Pare

Working...