Mouse Runs Through VR Maze

vr_mouse_setup

In an effort to determine the brain of animals during movement, scientists have built this contraption. It is a VR pod for a mouse. While we’re pretty sure/hopeful that none of you need a mouse VR system, we think the rig is interesting enough to stand on its own. It appears to be a convex mirror setup, projected in a dome. The controller is interesting in that it looks like a giant trackball hack. They’re using an optical mouse rigged to a ball floating on a cushion of air. This makes it much easier for the mouse to move. There’s a video of the whole thing in action after the break.

[youtube=http://www.youtube.com/watch?v=1DJOTEDBA2c]

Does that mouse look like he’s in control to you? he looks like he’s just instictually trying not to fall off of the giant ball. Notice that he never really stops or turns left. Just keeps running in that same general direction.

[via BoingBoing]

34 thoughts on “Mouse Runs Through VR Maze

  1. Hahahaha! Wow, they really didn’t do such a good job on the movement interface – the mouse seems very frustrated with how he’s positioned on the sphere. I do find the idea of an animal in a VR environment intriguing – does the animal navigate the environment(not here), if so, does the animal know it’s not real? In any case, I imagine the sphere needs to be a great deal larger in order for the subject to have a satisfying sense of balance and stabilitiy.

  2. is it me, or does it just look like the mouse is constantly running ‘right’ because it doesn’t want to be suspended by hooks over a rolling ball? I think a nascar game would have made for a better demo…
    but hmm, I wonder if there’s a better way to track motion, rather than a giant ball? some sort of floor that is an array of small balls with patterns on them, with a giant optical sensor below it (with the whole floor in view), looking up to determine the direction the balls are spinning? almost like a multi-touch setup, to determine the pace and direction of the user… I think that would be much easier for scale-up purposes. The smaller the balls and the closer packed they are, I imagine it will feel more natural to walk on… then again, I may just need to drink more coffee and let the brain churn some more… *sip

  3. A mouse is smarter than a cockroach. It’s not gonna feel comfortable and act normally when in equilibrium atop a ball, with its head held in a metal casing.

    What’s cool is that the VR rune the Quake engine. I can’t wait for the bionic mice/rat/cockroach/beetle/human deathmatch.

  4. @ chemical25 I think they do this normally run right,
    I mean isn’t the fool proof method of solving any maze to stick to the right? For instance if you put your right arm out and touched a wall and proceeded to walk forward eventually you would find the original exit or another one…? Of course the mouse has probes in its brain I can hear PETA now…… they could have made the graphics a little more realistic the mouse is prob like what the heck??

  5. @ chemical25

    I agree. It look interesting, but I’m far from convinced that the mouse wasn’t just running a large ball as opposed to running around in a VR environment ‘intelligently’.

    @ stinkymonkey

    you are making the assumption that no islands exist in the maze. the simulation wasn’t a maze either. i’m not convinced that the mouse was truly reacting to what it was seeing, there is only one instance that suggests this.

  6. The mouse continues to run to the next object that he can hide under. Obviously, the mouse isn’t going to just stop because, like someone already mentioned, it probably isn’t exactly comfortable on top of the ball or the rig attached to its head. It does seem to move through the room intelligently (room, not maze, it’s one large open space with low hanging objects to hide under).

  7. Fake… Show me the mouse turning left and right instead of running a circle as it feels it’s falling off a ball. It does not act like a mouse in a maze. it acts like a mouse freaking out.

  8. Seems a better approach would be to put the mouse inside a ball and then track it that way. The mouse would seem much more balanced I think.

    That being said, it would prevent the contraption they have attached to the mouse to take measurements.

  9. Thats what I was thinking. Put the mouse inside a hamster ball type device. Suspend the ball and have a projector mounted on the mouses’ back project an image onto the inside of the ball.

    Sounds fool proof to me.

  10. @turkeyhaXor They can’t do that because the head of the mouse will move and this wasn’t the purpose of the experiment. They had to attach the head to something (make it non-movable), to study the signals coming from single neurons. The mouse running in circles and all that has nothing to do with the experiment itself, since all they want to study is the relationship between the behaviour of the brain and the little guy’s running.

  11. The DLP projector is fail, because mice have a much higher flicker frequency than humans. I’m pretty sure the rainbow strobing of the projector is causing the mouse some serious “don’t wanna be here” type issues.

    It also looks like the ball could do with some damping, so it only moves when the mouse wants it to.

  12. first, mice don’t have higher flicker fusion frequencies than humans. Insects do, not other mammals. for example see (http://www.pnas.org/content/96/13/7553.full).

    Second, this maze isn’t even the one they used in the paper, so the video is just suppose to demonstrate the idea. In the paper they use a linear track, and show the mouse can run back and forth on the linear track and learns to do it better and better over time. And say what you want about the mouse’s movements on the ball.. if his brain shows the same kinds of signals as if he were navigating around a real environment, isn’t that the best evidence you have that the mouse is in virtual reality? and that’s what they found.
    see for example slashdot (http://bit.ly/fiCMM).
    or read the actual nature article.

  13. That paper doesn’t have anything to do with flicker fusion frequency.

    In any case, studies done on cats and rats have shown that they have two to three times higher fusion frequencies than humans. (Insects, due to their markedly different eyes, have much, much higher fusion frequencies still, but it’s still debated what they actually understand of this.) Human vision is a good “general purpose” tool, but these prey and carnivore animals have been in a “fast response” evolution war for millions of years.

  14. its not the primary focus but it does demonstrate it in figure 3
    http://www.pnas.org/content/96/13/7553/F3.expansion.html

    here is another paper..
    http://www.jneurosci.org/cgi/content/full/28/1/189

    again not direct, but its figure do suggest rod and cone sensitivity to gratings is dropping off when they go faster than 10Hz, such that it will be very small by 30Hz. I can’t find a direct reference on mouse flicker fusion, i suspect because the definition of flicker fusion is a behavioral one and people haven’t devised a behavioral test they are happy with for mice.

    i’d be happy to take a look at another reference, but I agree with you that cats do have better vision and higher flicker fusion, I haven’t read anything on rats, but what i’ve read about mice says they do not. A caveat to your evolutionary argument is that we are talking about lab mice strains which have not been under much selection pressure for vision. It is I think a valid criticism of this setup that they are only using vision given how poor lab mice see… but if it is sufficient to create place cells, I don’t see that that criticism is so severe.

  15. I think that mice normally follow a wall (hence the same direction). I think it would be more interesting to blind fold the mouse and put paddles on its whiskers… wait what is the point of this again?

Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.