How the brain encodes landmarks that help us navigate | Hacker News Reader

How the brain encodes landmarks that help us navigate

news.mit.edu

126 pointsprostoalex13 comments

Interesting. I've spent a lot of time "bushwhacking" off trail in our wilderness areas and one of the things I learned early on was to make a point to focus on landmarks, and more importantly, to turn around and look at where I'd just been so I could recall it on the way back because it's a completely different view.

It's a lot easier to travel off trail in areas with hills and hollows than flat lands with little variation in terrain and flora, and a compass makes a huge difference in that case, but even then stopping and turning around to look for something even a bit out of the ordinary can make a huge difference in rather you begin to panic over being lost or not.

Lol, I think you should rephrase yourself a little. I first read that as "I spent a lot of time bushwhacking off"

Reddit is that way.

> and more importantly, to turn around and look at where I'd just been so I could recall it on the way back because it's a completely different view.

This is thing I always forget. I have a pretty good sense of direction and spatial memory and knowing where I am and where I'm supposed to be going.

But I realized that if there's a road I'm often driving on back and forth, mentally, for me, the two directions might as well be two completely different roads. If I'm going in one direction, I have a really hard time knowing where I would be if I would turn around in the middle of it. I can't "connect" my two mental maps of going in each direction.

...compared to if it's a road I would walk along, because I turn around more often, look around more often, and then my mental map isn't two disconnected "tunnels", but a whole area.

This is also a good technique to not forget where you parked in a large parking lot.

Back when I worked in game development (level design), I found a really interesting pattern in World of Warcraft.

Basically, if you walked down a path, there would be something distinct every 20 seconds of walking. It might be a big rock on the side of the path, a bend, a little bridge over a creek, etc. Just something.

I think firstly, this served to help you feel like you were going the right way. A long and featureless path makes you wonder if you should turn around. One with features visible in the distance makes you want to keep going to see what's ahead.

Second (and relevant to the article), it helps you with navigation because it gives you easy to remember references.

Another comment to add, separate from the game thing. This reminds me of a patent a co-worker just put out: http://www.freepatentsonline.com/10578449.html

This is incredibly interesting, and something I wouldn't have thought much of before becoming a parent.

Since our son was able to speak, he was able to vocalize when we were near a family member's house (I saw this weekly as we drove from the San Francisco Peninsula up into the city). I was always curious what triggered him to know "we're nearby". We passed several unique structures (overpasses, apartment buildings, signs) but I was incredibly curious which ones he associated with being nearby this particular relative.

Eventually when he was a little older, and we asked, he was able to say a specific color building "told him" he was close. Super interesting. Child development and understanding in itself is so amazing to watch.

This is a cool example of science progressing in real time. I had heard previously that the hippocampus was related to navigation (the taxi driver study: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC18253/).

On the Wikipedia page for 'Retrosplenial cortex' (https://en.wikipedia.org/wiki/Retrosplenial_cortex#cite_note...), the intro text pulls on a source from 2009 to say, "its location close to visual areas and also to the hippocampal spatial/memory system suggest it may have a role in mediating between perceptual and memory functions."

And here there's new research making that link between perception and memory more clear!

One of my best friends was a long-term undergrad assistant on this project. It was always an eyebrow-raiser (and conversation starter!) when she declared that her research was developing "mouse VR."

Very cool to see it spotlighted on the MIT news site.

In both research and shipped products roboticists have been using landmarks for real-time simultaneous localization and mapping (SLAM):

http://ethaneade.org/eadeICCV2007.pdf

https://smartech.gatech.edu/bitstream/handle/1853/45226/Fact...

On a semi-related note did anyone watching the movie 1917 feel like the entire set was less than a mile wide? I was somewhat passively keeping track of distance covered and directions turned in my head and something just felt horribly off.

I swear they were on that truck for like 7 seconds and moved 100 feet.

Yeah, there's definitely a sort of implied (or attempted) transition in several scenes. From leaving the forest to reaching the cherry bushes, for example. The camerawork changes a bit and they subtlely crest a hill into the cherry bushes, if I recall correctly. I took these sorts of camera techniques to imply chronological progression.

I could discuss a few more but don't want to spoil. I had the pleasure of watching it twice in theatres and the cinematography, lighting, and editing just blew me away.

this "research" is the bio version of just running a benchmark on newer hardware and validating what was already known in hopes of stealing citations.

optically blocking neurons is such a cheap shot to validate old knowledge that was painstakingly researched with older tools. and they didn't even blast other random regions as a control.

Reply on news.ycombinator.com