Scientists Reconstruct Brains’ Visions Into Digital Video

**bihon** · 09-24-2011 04:00 AM

Just found this today.
Link to article here...

Scientists Reconstruct Brains' Visions Into Digital Video In Historic Experiment

What do you guys think?

**Oneironaut Zero** · 09-24-2011 04:05 AM

http://www.dreamviews.com/f45/breakt...mories-122116/

**ooflendoodle** · 09-24-2011 05:03 AM

Nothing is safe all is discovered run!!!!!11!1111!!

**Indeed** · 09-24-2011 05:11 AM

Man, if they had the database Watson used, but in video form, this would be crystal clear.

**Xei** · 09-25-2011 02:16 AM

What this technology actually is is a mystery to me... the article doesn't actually talk about how the computer knows which set of videos best match the brain scan, yet that's the crucial thing it sounds like they've done.

**Darkmatters** · 09-25-2011 02:18 AM

^^ Yup. I call foul. It's the new cold fusion.

**Xei** · 09-25-2011 02:36 AM

I'm not calling foul, it's just not a very helpful article.

And I imagine this technology is pretty limited, in theory, because looking at the mass flow of blood just won't give you the required fidelity. All they're getting right now is a few bits of information; it's not clear how much more you can get without going down to the neuronal level.

**Darkmatters** · 09-25-2011 02:40 AM

I just won't believe it till I hear something that makes actual sense.

**tommo** · 09-25-2011 03:36 PM

Originally Posted by Xei

I'm not calling foul, it's just not a very helpful article.

And I imagine this technology is pretty limited, in theory, because looking at the mass flow of blood just won't give you the required fidelity. All they're getting right now is a few bits of information; it's not clear how much more you can get without going down to the neuronal level.

This is basically what I said with the ones where the scientists used blood flow to see which pictures the subjects looked at/thought of.

As well as that you'd need a gigantic database for every possible thing people could see.

This does look a bit different, to get it really clear you'd still need a lot, just not every possible thing.

I also don't get how they figured out the response to the 18 million seconds of "palletes" either.

**Photolysis** · 09-25-2011 05:52 PM

It's interesting that they can use MRI to figure out what video clip is being looked at (once a database has been established), but it's somewhat limited. Still, I'm impressed they can determine that based on blood flow alone.

That said, it's very limited because it involves building up a huge database for every person. And it would appear that anyone who's worried about such devices being able to read your mind, that there's not too much to worry about because to enable that to happen in the first place would require you accurately telling them what you're experiencing and thinking. If you don't give answers, or give it deliberately wrong answers, then it can't figure it out.