Slashdot is powered by your submissions, so send in your scoop

 


Forgot your password?
Close
typodupeerror
×
Robotics Social Networks

Dogs Obey Commands Given by Social Robots (ieee.org) 76

Posted by msmash from the how-about-that dept.
Long time reader schwit1 shares a report: At the ACM/IEEE International Conference on Human-Robot Interaction (HRI 2020), researchers at Yale University's Social Robotics Lab led by Brian Scassellati presented a paper taking the first step towards determining whether dogs, which are incredibly good at understanding social behaviors in humans, see human-ish robots as agents -- or more specifically, whether dogs see robots more like humans (which they obey), or more like speaker systems (which they don't). The background research on dog-robot interaction that forms the basis for this work is incredibly interesting. The paper is absolutely worth reading in its entirety, but here are a few nuggets of prior work that should help you understand how dogs interact with non-human animated objects:

"Pongracz et. al tested whether dogs followed commands from their guardians with various levels of embodiment. The guardians may be present in the same room as the dogs (i.e., 3D condition), or interacted with the dogs via live-stream life-size interactive videos (i.e., 2D condition), or interacted with the dogs with only their voices came out of a loudspeaker (i.e., 0D condition). Dogs followed the commands most reliably in the 3D condition. They followed the commands least consistently in the 0D condition, and their performances were between 3D and 0D condition in the 2D condition.

Lakatos et. al conducted a study to test how dogs responded to the pointing cues given by a PeopleBot with customized arms. The PeopleBot either exhibited human-like behaviors or no social behaviors, depending on the condition. A dog participant observed the robot interacting with the guardian either socially or mechanically for six minutes in the interaction phase. The robot then delivered a food reward for the dog. In the subsequent testing phase, the robot pointed to one of the two buckets with hidden food rewards. In the testing phase, dogs performed better in the condition with a social robot than with a nonsocial robot. However, no evidence suggested the mean performance with the social robot was significantly different from 50 percent, which is the chance level in two-choice tasks. Therefore, the dogs did not consistently follow the pointing cues provided by the social robot, even though dogs in general follow human pointing cues well."
This discussion has been archived. No new comments can be posted.

Dogs Obey Commands Given by Social Robots More

Dogs Obey Commands Given by Social Robots

Comments Filter:

  • Well duh (Score:4, Insightful)

    by Calydor ( 739835 ) on Friday May 15, 2020 @02:12AM (#60062324)

    Dogs grow up around humans literally from the moment they're born. Their mother is seen to obey what humans do and usually show adoration for the humans. The puppies mimic this behavior.

    If you take a dog that has basically never had any interaction with humans it is not going to obey any cues, even if the human dresses up like a dog and tries to mimic dog cues. I suspect the same is the case here - the dogs look at the robots and go, "Huh, that's a weird thing." They've seen and heard us arguing with the TV, with the radio, with disembodied voices on the telephone etc. Why should they see a robot as any different from so many other weird things humans have?

    • Re: (Score:2)

      by AmiMoJo ( 196126 )

      Dogs are pack animals, they have a natural instinct to follow the leader of the pack which in a domestic setting is their owner.

      • Dogs are pack animals, they have a natural instinct to follow the leader of the pack which in a domestic setting is their owner.

        Maybe, you are assuming the human owner is the pack leader.

        Here's a test. Does the dog listen to you when you do NOT have food, treats or a toy in your hand? Yes, leader. No, not leader.

        • Of course it does... Does yours not?!

          • Re: (Score:2)

            by drnb ( 2434720 )

            Of course it does... Does yours not?!

            I've only needed the occasional verbal praise once trained. In the beginning of training it was a combination of physical affection, rubbing the neck, scratching behind the ears, while offering verbal praise. This associates the verbal praise with affection.

            Training, play and feeding times were all separate. Treats were random events that never occurred during training or at the kitchen table.

            • I personally went with the rigid treat approach: pup gets a treat every night at 7.

              I think I misunderstood the point you were trying to make. I had thought you were saying a dog was merely some kind of treat-seeking automaton... Iâ(TM)ve never needed to carrots to get my dog to do what I wanted

              • Re: (Score:2)

                by drnb ( 2434720 )
                Awkward wording, my apologies. I was trying to make the point that if you *need* to use treats/toys, if your commands are generally ignored without them, then you are not recognized as a leader and compliance with commands is optional.

    • Why should they see a robot as any different from so many other weird things humans have?

      Because it gives a reward of food.

      I am a little surprised there is a direct "social" effect.

      For example, if pushing through a doggie door will get them access to food, they can certainly learn to do that. If ringing a bell is followed by food, they learn an autonomic response to that (Pavlov's dog). So, clearly dogs don't need social interaction with humans to be part of reward-seeking behavior.

    • Re: (Score:2)

      by Falos ( 2905315 )

      Domesticated species have been around long enough to have inclinations in their DNA. Note the distinction - DNA doesn't directly tell a bird how to fly, but it can blueprint for brain wiring pre-shaped for that purpose. Domestic species have increased their ability to read humans, as well as be read (eg expressive eye muscles only in domestics), something present in a no-interaction specimen.

      Domestication dates back far enough that we believe even humans were affected [wikipedia.org], "domesticated" say the clickbaits. We

  • and an intelligent Tennis ball machine who raised them - and spat cats at the police when they tried to enter.

  • Wait, what? (Score:5, Insightful)

    by 93 Escort Wagon ( 326346 ) on Friday May 15, 2020 @03:31AM (#60062438)

    Title: Dogs obey commands given by humanoid robots

    Conclusion in the summary: Therefore, the dogs did not consistently follow the pointing cues provided by the social robot, even though dogs in general follow human pointing cues well.

    Why do we keep seeing articles posted here where the title claims something unsupported by the article underneath it? Another recent example was the “stupid Americans are killing millions of bees because of murder hornets” post.

    • Re: (Score:2)

      by AmiMoJo ( 196126 )

      I wonder if it's because they are not just following the pointing when a human does it, they are following the human's eyeline and other cues.

      I don't know about dogs but cats don't understand pointing. They look at your hand and wonder why it's stuck out. Maybe dogs learn to go where you point, maybe they just look where you seem to be focusing your interest.

      Robots are usually very bad at the kind of subtle cues that humans give off.

      • Or Cats don't want us to know they understand pointing. A long running scam to avoid the types of jobs we give to dogs, like guard dogs, sniffer dogs, etc.. ;)

        • Re: (Score:2)

          by AmiMoJo ( 196126 )

          I suspect that may be it. The cat just watches you throw a ball and then looks at you like "why did you throw your ball away?"

          When they are young they might play a bit, but they quickly learn that if they just lie there the string soon comes to them and there is no need to chase it around. Let the human do all the work.

          • The problem is that, usually, cats do not see a ball as a prey target. The dot from a laser pointer will get them motivated quickly.

            • Laser pointer can work with a dog too. One of mine loved it and would chase til exhaustion. The other I think was blind to it. Got very upset that the other dog was having fun but couldn't figure out why and never responded to it at all despite being super playful in other ways.

        • Cats look at your finger. Dogs look where you're pointing (and looking).

          • Nope. That's just something you heard a human say, and so you repeated it. Human hear, human say.

            • I also heard a human (you) say that they don't.

              I've tried it. Cats really do just look at the finger.

              • That's why I explain things, but don't provide links.

                Don't believe it because I said it. If you're smart enough to consider the content, do so.

                Believe nothing, including your own conclusions. Keep an open mind. Be used to operating on your best guess with the knowledge you were able to collect in the time you had available.

                You tried "it," whatever "it" is, and you got results consistent with the reported results. But you confused correlation with cause; that's bundled into the word "just" in your statement.

                • I don't think a cat knows how to deal with a bear other than to hiss and then run. I think a bear will treat a housecat vastly differently than it treats a human, so expecting a cat to follow advice formulated for humans is a bit stupid on multiple levels.

                  • No hiss. Definitely no hiss.

                    And

                    expecting a cat to follow advice formulated for humans is a bit stupid on multiple levels.

                    Nobody said that, so yeah... on multiple levels.

      • Cats can certainly understand pointing with proper training. Start pointing things right at the tip of your finger, ex: treats, then over time increase the distance, at some point they "click" and understand pointing. It helps if you look at what you're pointing, too. They do needs a lot more training than dogs, only patient people can do it. Always look at it from the cat's point of view: what's in it for him/her?

        • The first problem people have with this is that cats don't put the same social significance on looking at things that humans do. Cats are very careful not to make eye contact unless it is socially appropriate. Humans generally can't comprehend that; they'll try to look the cat in the eye to communicate, not realizing that the cat is already carefully watching them while its eyes are focused somewhere else! They don't understand why the cat "can't" learn to do it the human way, while the cat hasn't even seen

      • Dogs are pack animals in the wild, and so taking directions from others is somewhat instinctual. They will also "point", even in non hunting dogs this behavior can be seen. The nose and body points towards the prey, and the tail is up possibly as a signal to others. Cats are mostly solitary (well, small cats, not lions) and don't have instincts to work as a team.

      • Re: (Score:2)

        by ceoyoyo ( 59147 )

        Maybe your cat just isn't interested in whatever you're pointing at. My friend has two cats, and they will absolutely follow a pointed finger to find a toy, *if* they want to play.

        A long time ago I dated a girl who had an absolutely evil cat. If you shone a laser pointer across the room it would sit and watch for a while, working out where the light was coming from. Then it would attack you.

    • Inflammatory clickbait headlines on SLASHDOT?! No way! Not in 2020!
      Seriously though, this place needs new editors.

    • Conclusion: most dog owners are indistinguishable from robots.

    • I thought the same as you, but the reference in the summary was about an earlier study with a nonhumanoid robot. The study in TFA refers to is not about gestures, it is about a vocal command (sit) given by a humanoid robot. The dogs obeyed 60% of the time, as opposed to 20% of the time with only a speaker.

      So headline ok, slashdot summary is still garbage.

    • Why do we keep seeing articles posted here where the title claims something unsupported by the article underneath it?

      Come on, how are they expected to understand that "However, no evidence suggested the mean performance with the social robot was significantly different from 50 percent, which is the chance level in two-choice tasks" means that it failed?

      They're just editors, and 50% isn't 0%, so how would they know? You can't expect so much from people; how could editors hope to comprehend words?

    • Why do we keep seeing articles posted here where the title claims something unsupported by the article underneath it?

      History will call it ...a movement that lead to eventually becoming Trumpist Dogma: Your projection of whatever reality is, is the reality you, and everyone else, live.

      Its an even more perverted form of the Socratic method -- instead of even asking disingenuous yes/no questions, we are left with rhetorical statements of fact, asking a serious of disingenuous questions.

      It would be farcical if not for the fact of its current enjoyed progressive use.

    • dogs obeys robots order that's new think our company also deal with new idea AICRA

  • Eh. (Score:3)

    by SCVonSteroids ( 2816091 ) on Friday May 15, 2020 @03:52AM (#60062464)

    Robot: Sit, dog!
    Dog: No you sit, dog.

  • That one for a cellular company where an old couple keeps track of the dog with there phone, The woman then tells the dog to get off the couch, which the dog does. Oh well, another illusion shattered.
    • That's the dogs "oh shit I didn't hear them come in, 'I'm not on the couch' jump to the floor quickly and look innocent" response.

    • Re: (Score:2)

      by drnb ( 2434720 )
      You can probably train the dog to get off the couch when the cell phones says too.

      Oh and dogs can recognize family over the phone. After being away for almost two weeks I called home and at the end of the conversation my mom put the phone near the dog's ear so I could say hi. The dog went nuts running between front door and back door looking for me.

  • So do humans, once we start adding weapons to our social robots.

  • Why follow an order of a being without any arm to give you a treat?

  • If you train them, dogs will obey lights, so that a taught dog can follow a robots instructions doing the thing the dog was trained to react to is kind of obvious?

    • They also recognize voices, so it is hard to say that they even know a "robot" is involved when their owners voice comes out of it.

      It may be that dogs have learned that televisions aren't real, so they distrust the voice when it is combined with the video image, but with the robot they can't figure out where the voice is coming from, maybe their owner is inside or behind the robot, and therefor nearby, so they might need to obey.

      They seem a long way from discovering how the dog feels about the situation, ev

  • At least they used a dog to experiment first.

  • The robot then delivered a food reward for the dog.

    My one year old son drops food on the ground for my dog to eat. My dog now listens to my one year old son. It's really not that complicated.

    • Re: (Score:2)

      by ceoyoyo ( 59147 )

      One of my cats growing up did the same thing. And I knew someone who had a parrot that did too.

  • Is anyone surprised by this? It's all tinkering at this point.
    • Esp NZ Sheepdogs. The dogs look into your eyes and face as well for clues, tone of voice, and aim to please. They get clues from whistles -no hands needed - handy when you are on a motorbike. In NZ they love attaching TV antenna to the dogs head and have an oversized RC radio control to move the dogs about. They can even read the finger movements. But some cheat by adding yellow and blue spots to the front of the control box. Then whistle when the dog makes a mistake. Tourists are totally conned.

  • Dogs listen to parrots too.. (Score:5, Interesting)

    by quetwo ( 1203948 ) on Friday May 15, 2020 @07:58AM (#60062900) Homepage

    When I was growing up, we had an African Grey parrot and a Cockatoo. They would regularly give instructions to our dogs, who would follow them gleefully. Surprisingly, the dogs would also respond to praise and reprimand from the birds too. "Sammy, go get the ball!" and the dog would get the ball and drop it front of the bird's cage. "Sammy, you're baaaadddd," and the dog would go away, cowering in the corner.

  • Bad testing methods (Score:3)

    by DontBeAMoran ( 4843879 ) on Friday May 15, 2020 @08:11AM (#60062928)

    The guardians may be present in the same room as the dogs (i.e., 3D condition), or interacted with the dogs via live-stream life-size interactive videos (i.e., 2D condition), or interacted with the dogs with only their voices came out of a loudspeaker (i.e., 0D condition).

    Okay so, no 1D condition? That's just lazy.

    • They couldn't figure out if an audio track was a variable or not. They're just researchers, not hard scientists, give them a break.

  • Sadly, there's plenty of people who obey commands given by social robots as well... Just look at Democrats.

  • Dogs are pretty bright. They can understand lot of things. They are born with a set of built in gestures they all understand and use -- pointing, begging, etc. So it's not surprising that they understand gestures.

    As to "obeying". Not my experience. Mostly, dogs listen, watch, figure out what is desired. Then they do whatever they damn well please.

    They are a lot like people.

    • Better than cats. They listen, watch, figure out what is desired, and then do the exact opposite of that while giving you a shit-eating inquisitive look just to let you know where you're place is at.
      • Which is similar behavior to child with very severe developmental disorders.
        What I don't understand is why in cats, people ascribe this behavior to some kind of hyperintelligence.
        The theory that cats are so brilliant that they know how to hide all visible signs of higher intelligence really does make my head hurt.

Slashdot Top Deals

Some people manage by the book, even though they don't know who wrote the book or even what book.

Close