Author Options:

Could dogs and cats communicate with people using this eye tracking method? Answered


This post is a few days old.   I posted a comment asking the same question on HAD, but I thought it may get more exposure here.

If you've ever noticed, dogs pupils really reflect light, even visible light like from a cars headlights.
I would think this system would be able to track a dog's eye movement as well.
This would need to start out much simpler, to teach the animal to use the system.   Perhaps give him 2 choices, in the form of recognizable icons.   I am imagining a picture of a steak and a picture of a dog going for a walk.   Eventually the dog might learn to recognize the symbols.   Once this happens, the interface could be expanded to more choices.   I dont know if it would be possible to teach english ( or any other spoken/written language) but maybe.



8 years ago

You can teach a dog visual cues through association.

For instance, a dog will learn that you picking up the leash means going for a walk.  If the dog enjoys going for a walk, then it will go to the door.  If the dog hates going for a walk, then it will hide.  The leash could also come to mean going to the vet or going for a car ride which might elicit different responses.

To condition the dog to recognize a cue card in place of physically picking up a leash, you would need to pair the card with the leash.  For instance, you show the card to the dog and then pick up the leash and go for a walk.  Eventually, you will only need to show the card to elicit the original response of picking up the leash.  But honestly, you could pair picking up the leash with any symbol - you could use a star or a heart which would be easier for a dog to differentiate from another image.

Dogs are not capable of understanding English though.  Their brains are different, and they may only understand up to 70 words.  It's impossible to say if it's the word itself or the sound of the word.  More than likely, it's just the sound of the word.  Dogs learn their names faster if given short, clippy-sounding names like Scout rather than Benjamin.

I would recommend reading For the Love of a Dog by Dr. Patricia McConnell to better understand the psychology of dogs, and I would recommend trying out this theory with a dog rather than a cat because humans and dogs communicate more easily with one another after a very long time of living beside one another.  And you might want to brush up on some behavioral psychology generally.

And then there's the issue of breed.  A herding dog that relies heavily on visual cues through selective breeding, is very intelligent, and motivated to please its owner (e.g., a border collie) would be best.  I don't think you'll get much from a shih tzu - they're cute, friendly, and silly but not the smartest or easiest dog to teach.

And yes, people have tried to teach dogs ways to communicate.  People have tried to teach themselves how to communicate with dogs.  Dogs learning people communication.  People learning dog communication.  I don't think eye tracking would be very practical, but it's fun to see how much dogs can learn.  It's fun to see how much people can learn as well.

Best of luck on your behavioral studies!


8 years ago

I would say yes me and my cats and dogs kind of know what each other wants I can tell if they are angry at me happy or sad because they look at me a different way for each one


8 years ago

You might want to do a literature search. People have certainly experimented with other ways of letting animals indicate what they want via pedal switches and the like. There are real limits, and I don't think making the mechanism portable would extend them much.

And pictures aren't all that recognizable for most animals. If it doesn't smell right, and it doesn't move right, it isn't the thing. Strongly differentiated abstract symbols, arbitrarily chosen, might work better.

Remember that many dogs don't have great vision, and don't see the same range of colors that we do. Cats have better vision but still have the color limitation.

It could be an interesting experiment to try, but I wouldn't hold out much hope for much success.


8 years ago

.  It should be possible to train a dog to use eye tracking for simple communication, but I don't see it opening up new avenues of thought. Almost certainly not something that would allow them to use language any better than they can already (ie, close to not at all). More like using a foot switch or bark activated switch.
.  Apes and cetaceans might be able to make better use of eye tracking for complex communication.
.  But I'm no expert.