Neurons subtract images and use the differences
Scientists thought that your eyes operated like video cameras, transmitting images in their entirety. The brain might not process all of the detail, but the images were nonetheless there. Now, new research from the Ruhr-Universität in Germany, has found that our neurons compress data by 'subtracting' or removing parts of visual scenes that haven't changed.
The work, published in the journal Cerebral Cortex, displayed two versions of two different images (a landscape scene and a building). One version of the image was a photograph. In the other image, the researchers digitally removed some of the vertical or horizontal lines from the image, making a less detailed, less data-rich image. When the researchers displayed the images in rapid succession, with a gap of 30 milliseconds or less between them, the neurons transmitted complete information. But when the gap was longer (around 100 milliseconds), the neurons only transmitted the information that was new or different in the second image. This makes potentially interesting and useful areas of a scene easier to detect by our brain and shows how past events are continually influencing what we see.
Source:
http://aktuell.ruhr-uni-bochum.de/pm2013/pm00332.html.en
Journal article: Primary visual cortex represents the difference between past and present, Cerebral Cortex, 2013. DOI: 10.1093/cercor/bht318
Story via Neuroscience Research Techniques
Image via Je
Vision and thought are connected ...you're perceiving prioritizing and reacting to visual stimulus all at once.time lapses may be unimportant to cognitive vision if it was you would account for it..we did tests of visual acuity in 3&400ths of a second and you learn to process as required at miaculous cognitive speed.. sports are great examples of this tennis squash racquetball hockey etc etc.. miracles at work!!
ReplyDeleteIs true we do learn from the past and neurons fill the gaps when something is missing from the original "vision". Also experience helps you anticipate movements especially when it comes to sports.
ReplyDeleteI think we've known for at least 15 years that neurons do a lot of preprocessing, causing the eye to track patterns and then sending information on detected edges and features. Also, anyone who has played with domesticated birds can tell you they have a change-sensitive vision system when they're quasi-sleeping.
ReplyDeleteYes but this study shows that our brain runs in .mkv ;)
ReplyDeleteJohn Bump yes this is why change of speed or spin and element of suprise is so important in sport..we are so visually intuitive that if you become predictable your dead..lol
ReplyDeleteIt's probably also part of why we like fires, fishtanks, and the sea: something constantly moving plays with parts of our brains.
ReplyDeleteOf course..your brain does not alert you to what you have already seen - knowing that has been around for quite sometime.
ReplyDeleteIt's a constant "learn-and-store" process.
Corina Marinescu there is an alchemy of old and new knowledge that can even be combined with imagination and creativity that can be incorporated .. we see animals and faces in clouds or objects. Use previous experience to enhance our visual journey!!
ReplyDeleteYes pareidolia is quite interesting :)
ReplyDeleteA lot of the radio transmitters we use work on a similar principle. Each unit transmits only when there is a change to the readings they are monitoring. This saves a lot of energy at the transmitter and processing time at the receiver. I would suspect our vision developed for similar reasons. Saving energy and allow more time to process the affect of any differences, which are more critical to our survival.
ReplyDeleteThe Science Channel has a great series on perception. Check out Your Bleeped Up Brain. It's "eye opening."
ReplyDeleteThat's an interesting aspect Kenny Laidlaw.
ReplyDeleteOur brains record all incoming data. The conscious mind receives new info while reoccurring info is routed to the subconscious. That wall keeps us from drowning in too much information but it also separates us from quite a few of the brain's abilities. There is a single location in our brains in which all info goes to before it's routed to any other part of the brain. That one location is the place where awareness grows.
Eric Pike thanks for the reference.
ReplyDeletethere's something like "feed before"....(as opposed to "feed-back" mechanism).
ReplyDeletehumm....
I wonder how far the parallel between human designed video compression algorithms and biological visual systems goes. For example, does the prefrontal cortex (or other parts of the brain) understand inter-frame movement of objects? Is there even some form of intra-frame compression? (Here's a quick overview of MPEG compression for reference http://www.abc.net.au/http/sfist/mpeg1.htm) Looking at the information processing schemes of biological systems is an intriguing prospect. Once understood, I have a feeling we'll be marveling at both 'brilliance and the 'stupidity' of them much as we do with the mechanical characteristics of biological systems.
ReplyDeletevictor reus -- feedforward is a term used in process control and in the study of regulation of enzyme cascades, where input process rates alter output process rates.
ReplyDeleteThis study shows how our brains work kinda like a video encryption Sean Walker . This effect also seem to apply to biographical memory. For example, we seem to recall the present and apply a change algorithm to reconstruct the past.
ReplyDeletethanks John Bump !
ReplyDeletethe meaning for 'feedbefore' was, lets say, metaphorical....
Sean Walker very interesting article on how we perceive process and articulate visually. One point in there is incorrect we intially see things as a blur at high speeds but the eye can be trained to recognize and process at speeds beyond what is perceived to be normal we studied this visual acuity with the best athletes in.the world. It's miraculous. And if you change the chemistry of processing this observational skill it can be slowed as with alcohol or even lsd where you get traces( so they tell me) frame by frame if you will...lol
ReplyDeleteKeith Robinson I didn't understand what you meant by 'articulating visually'. In any case, due to my generously misspent youth I can confirm the LSD effect on one's vision. And I thought being a former miscreant would have no practical benefit in my life - just goes to show, you never know. :)
ReplyDeleteCorina Marinescu I thought the article said something more definite about visual compression and offered specific evidence for it - ie the plus-minus coding observations of the neurons. I got the 'biographical' gist also, but it seemed more speculative or at least not fleshed out. I also have the impression that past experience projection in visual processing is likely limited more to recent-past visual data. I would imagine that generous use of past visual experience information would not be possible because of the logical problem of mapping large spaces of visual information for 'lookup' on the fly.
ReplyDeleteThe biographical memory is not part of this study...Also my fusimotors are telling me to pay attention to driving and to stop comment on G+ =)
ReplyDeleteCome to think of it though, I guess something like a filtering approach in which significant images are stored in the brain for rapid look up in visual processing could undo the problem I suggested. Significant images could be visual patterns that we've been presented with more often or that are emotionally more impactful to us. For example, loved ones.
ReplyDeleteMost technology is based on how things are done in nature, with or without intending to. You can't really beat millions/billions of years or trial and error. The best, or most adaptable solution being the one to carry on and getting the chance to develop further.
ReplyDeleteWhat about when you blink? Everything changes as it goes dark for a fraction of a second. Your brain merely maintains the last image during this. How does the process compensate for this? In theory, everything has changed during that blink but in reality nothing could have changed.
To give you an idea what I mean. Watch a spinning disc or wheel and blink rapidly. You should see the image stop and start, like in a zoetrope.
Kenny Laidlaw Although your point was not that all technologies have strong parallels in biology, I'm not convinced that most do. The wheel for example. There may be some vague examples of it, but it certainly isn't a common macro level biological pattern. Other technologies that come readily to mind also don't appear to me to have any parallels in biology: paper, HTML, ball bearings... I may be just unaware of the parallels for these examples though.
ReplyDeleteI can think of one that do have clear parallels too, but they came less rapidly to me - the piston, camera, levers.
Is this an algebraic word problem?
ReplyDeleteSean Walker Maybe I should rephrase it. "Most technology" was more to how things are designed and not the actual technology or how we program it to work. HTML (although DNA could be seen as along the same lines).
ReplyDeleteI was thinking more along the lines of how a camera lens is designed along the same lines as your eye, the lens to focus, the shutter working as your iris and film as the photoreceptors on the back of your eye.
You mentioned ball bearings. These can be made by dropping the molten metal and letting nature do the rest. Making the ball bearing into a sphere using surface tension. Not really designed by nature, but the most energy efficient shape in nature.
A wheel would be a less efficient design of a ball. The ball having full 3D movement, just ask James Dyson. Which leads to hoovers now designed based on nature.
I'm not saying everything in nature is the best solution, or that man couldn't make things better but nature does give you a good base to begin with. Man can remove any "vestiges", whilst nature tends to be stuck with these.
OK you two...you moved this subject to a tech floor..no bueno =)
ReplyDeleteThis is all neuro =D
Ok mom Sis.
ReplyDeleteNeuro question. Are some forms of autism caused when the processing fails and the brain is swamped by all the data and not filtered for just the differences.
Mom? Thanks a lot for making me feel ancient Kenny Laidlaw =)
ReplyDeleteCorina Marinescu Very true, I am definitely the older of the two of us.
ReplyDelete