A headset that reads your brainwaves | Tan Le

Prevoditelj: Mislav Ante Omazić - EFZG Recezent: Predrag Pale

Up until now, our communication with machines

Sve do sada je, naša komunikacija sa strojevima

has always been limited

uvijek bila ograničena

to conscious and direct forms.

na svjesne i izravne postupke.

Whether it's something simple

Bez obzira radi li se o nečem jednostavnom,

like turning on the lights with a switch,

poput paljenja svjetla prekidačem,

or even as complex as programming robotics,

ili o nečem složenom poput programiranja robota,

we have always had to give a command to a machine,

oduvijek smo trebali dati naredbu stroju,

or even a series of commands,

ili čak niz naredbi,

in order for it to do something for us.

kako bi on nešto napravio za nas.

Communication between people, on the other hand,

Komunikacija između ljudi, s druge strane,

is far more complex and a lot more interesting

je mnogo složenija i puno zanimljivija,

because we take into account

jer uzimamo u obzir

so much more than what is explicitly expressed.

puno više od onoga što se izričito iskazuje.

We observe facial expressions, body language,

Promatramo izraz lica, govor tijela,

and we can intuit feelings and emotions

i možemo osjetiti emocije

from our dialogue with one another.

u razgovoru s drugima.

This actually forms a large part

To je zapravo veliki dio

of our decision-making process.

našeg procesa donošenja odluka.

Our vision is to introduce

Naša vizija je uvesti

this whole new realm of human interaction

tu potpuno novu dimenziju ljudske komunikacije

into human-computer interaction

u interakciju između čovjeka i računala,

so that computers can understand

kako bi računala mogla razumijeti

not only what you direct it to do,

ne samo ono što im naredite da naprave,

but it can also respond

već kako bi mogla reagirati

to your facial expressions

na vaš izraz lica

and emotional experiences.

i emocionalna iskustva.

And what better way to do this

A koji je bolji način za to napraviti

than by interpreting the signals

od interpretacije signala

naturally produced by our brain,

koje prirodno proizvodi naš mozak,

our center for control and experience.

naš centar za kontrolu i iskustvo.

Well, it sounds like a pretty good idea,

Zvuči kao prilično dobra ideja,

but this task, as Bruno mentioned,

ali taj zadatak, kao što je Bruno spomenuo,

isn't an easy one for two main reasons:

nije jednostavan iz dva glavna razloga:

First, the detection algorithms.

Prvo, algoritmi za detekciju.

Our brain is made up of

Naš mozak je napravljen od

billions of active neurons,

milijardi aktivnih neurona

around 170,000 km

i oko 170,000 km

of combined axon length.

ukupne dužine aksona.

When these neurons interact,

Kada su ti neuroni u interakciji,

the chemical reaction emits an electrical impulse,

kemijska reakcija emitira električne impulse

which can be measured.

koji se mogu mjeriti.

The majority of our functional brain

Većina naših moždanih funkcija

is distributed over

je raspodijeljena po

the outer surface layer of the brain,

vanjskoj površini mozga.

and to increase the area that's available for mental capacity,

A kako bi se povećalo područje koje je dostupno za mentalne aktivnosti,

the brain surface is highly folded.

površina mozga je jako naborana.

Now this cortical folding

Sva ta kortikalna pregibanja

presents a significant challenge

predstavljaju značajan izazov

for interpreting surface electrical impulses.

za interpretaciju površinskih električnih impulsa.

Each individual's cortex

Korteks svake osobe

is folded differently,

je naboran drugačije,

very much like a fingerprint.

nešto poput otiska prsta.

So even though a signal

Pa premda signali

may come from the same functional part of the brain,

možda dolaze iz istog funkcijskog dijela mozga,

by the time the structure has been folded,

ali zbog različitog nabiranja mozga

its physical location

njihova fizička lokacija

is very different between individuals,

značajno varira među pojedinicima,

even identical twins.

čak i među identičnim blizanicima.

There is no longer any consistency

Tako nema nikakve konzistentnosti

in the surface signals.

u površinskim signalima.

Our breakthrough was to create an algorithm

Naše otkriće je što smo napravili algoritam

that unfolds the cortex,

koji je "izravnao" korteks,

so that we can map the signals

kako bismo mogli mapirati signale

closer to its source,

bliže njihovu izvoru,

and therefore making it capable of working across a mass population.

što nas čini sposobnim za rad s masama.

The second challenge

Drugi izazov

is the actual device for observing brainwaves.

jest sam uređaj za promatranje moždanih valova.

EEG measurements typically involve

EEG mjerenja tipično rade

a hairnet with an array of sensors,

s mrežom senzora na glavi,

like the one that you can see here in the photo.

poput ove koju možete vidjeti na slici.

A technician will put the electrodes

Tehničar će staviti elektrode

onto the scalp

izravno na glavu

using a conductive gel or paste

koristeći provodivi gel ili pastu

and usually after a procedure of preparing the scalp

i obično nakon pripreme glave

by light abrasion.

laganim struganjem kože.

Now this is quite time consuming

To sve zahtjeva dosta vremena

and isn't the most comfortable process.

i nije najudobniji proces.

And on top of that, these systems

I pored svega, ti sustavi

actually cost in the tens of thousands of dollars.

koštaju desetine tisuća dolara.

So with that, I'd like to invite onstage

Imajući to na umu, željela bih pozvati

Evan Grant, who is one of last year's speakers,

Evana Granta, koji je jedan o prošlogodišnjih govornika

who's kindly agreed

koji je ljubazno pristao

to help me to demonstrate

pomoći mi demonstrirati

what we've been able to develop.

ono što smo uspjeli razviti.

(Applause)

(Pljesak)

So the device that you see

Dakle, uređaj koji vidite

is a 14-channel, high-fidelity

jest 14-kanalni, jako precizni

EEG acquisition system.

EEG sustav za mjerenje.

It doesn't require any scalp preparation,

Ne zahtjeva nikakvu pripremu glave,

no conductive gel or paste.

nikakav provodivi gel ili pastu.

It only takes a few minutes to put on

Treba mu samo nekoliko minuta da se postavi

and for the signals to settle.

i da se signali smire.

It's also wireless,

Također je bežičan,

so it gives you the freedom to move around.

tako da vam daje slobodu kretanja.

And compared to the tens of thousands of dollars

I u usporedbi s desetinama tisuća dolara

for a traditional EEG system,

za tradicionalni EEG sustav,

this headset only costs

ovaj uređaj košta samo

a few hundred dollars.

nekoliko stotina dolara.

Now on to the detection algorithms.

A sada nešto o algoritmu za prepoznavanje signala.

So facial expressions --

Dakle izrazi lica --

as I mentioned before in emotional experiences --

bitni za spoznaju o emocionalnim iskustvima --

are actually designed to work out of the box

su dizajnirani tako da funkcionira bez podešavanja

with some sensitivity adjustments

a osjetljivost se može fino prilagoditi

available for personalization.

u ovisnosti o korisniku.

But with the limited time we have available,

Ali s obzirom da nam je vrijeme ograničeno,

I'd like to show you the cognitive suite,

željela bih vam pokazati kognitivnu aplikaciju,

which is the ability for you

koja omogućuje

to basically move virtual objects with your mind.

da umom pomičete virtualne objekte.

Now, Evan is new to this system,

Evan nije prije koristio ovaj sustav,

so what we have to do first

zato prvo moramo

is create a new profile for him.

napraviti novi profil za njega.

He's obviously not Joanne -- so we'll "add user."

On očito nije Joanne -- dakle "dodati ćemo korisnika".

Evan. Okay.

Evan. OK.

So the first thing we need to do with the cognitive suite

Dakle prva stvar koju moramo napraviti u kognitivnoj aplikaciji

is to start with training

jest započeti s podešavanjem

a neutral signal.

na neutralni signal.

With neutral, there's nothing in particular

Na neutralnom, Evan ne treba ništa posebno

that Evan needs to do.

napraviti.

He just hangs out. He's relaxed.

Miran je. Opušten.

And the idea is to establish a baseline

I ideja je da se formira baza

or normal state for his brain,

ili normalno stanje njegova mozga,

because every brain is different.

jer je svaki mozak drugačiji.

It takes eight seconds to do this,

Potrebno je osam sekundi za ovo.

and now that that's done,

I sada kada je to napravljeno,

we can choose a movement-based action.

možemo izabrati akciju vezanu za pokret.

So Evan, choose something

Dakle Evane izaberi nešto

that you can visualize clearly in your mind.

što možeš jasno vizualizirati u svom umu.

Evan Grant: Let's do "pull."

Evan Grant: "Idemo pokušati "povući"."

Tan Le: Okay, so let's choose "pull."

Tan Le: "OK. Izaberimo "povući"."

So the idea here now

Dakle ideja je

is that Evan needs to

da Evan mora

imagine the object coming forward

zamisliti objekt koji dolazi prema naprijed

into the screen,

na ekran.

and there's a progress bar that will scroll across the screen

Ovdje je pokazivač napretka koji klizi preko ekrana

while he's doing that.

dok on to radi.

The first time, nothing will happen,

Prvi puta, ništa se neće dogoditi,

because the system has no idea how he thinks about "pull."

jer sustav nema pojma na koji način on zamišlja "povuci".

But maintain that thought

Ali zadrži tu misao

for the entire duration of the eight seconds.

svih osam sekundi trajanja.

So: one, two, three, go.

Dakle: jedan, dva, tri, počni.

Okay.

OK.

So once we accept this,

Kako smo jednom unijeli naredbu,

the cube is live.

kocka postaje živa.

So let's see if Evan

Pogledajmo može li Evan

can actually try and imagine pulling.

stvarno zamisliti "povlačenje".

Ah, good job!

Ah, dobar posao!

(Applause)

(Pljesak)

That's really amazing.

To je prilično zadivljujuće.

(Applause)

(Pljesak)

So we have a little bit of time available,

Imamo još malo vremena na raspolaganju,

so I'm going to ask Evan

pa ću zamoliti Evana

to do a really difficult task.

da izvede stvarno težak zadatak.

And this one is difficult

A ovaj je težak

because it's all about being able to visualize something

jer treba zamisliti nešto

that doesn't exist in our physical world.

što ne postoji u stvarnom svijetu.

This is "disappear."

To je naredba "nestani".

So what you want to do -- at least with movement-based actions,

Ono što želimo -- barem vezano s akcijom kretanja

we do that all the time, so you can visualize it.

to radimo stalno, tako da je možemo vizualizirati.

But with "disappear," there's really no analogies --

Ali s "nestani", ne postoje stvarne analogije.

so Evan, what you want to do here

Evane, zato sada želimo da

is to imagine the cube slowly fading out, okay.

zamisliš kocku kako polako nestaje, OK.

Same sort of drill. So: one, two, three, go.

Isti postupak. Dakle: jedan, dva, tri, počni.

Okay. Let's try that.

OK. Idemo pokušati to.

Oh, my goodness. He's just too good.

Oh, moj Bože. On je jednostavno predobar.

Let's try that again.

Pokušajmo ponovno.

EG: Losing concentration.

EG: "Gubim koncentraciju."

(Laughter)

(Smijeh)

TL: But we can see that it actually works,

TL: "Ali možemo vidjeti da to stvarno radi,

even though you can only hold it

premda se ne možeš koncentrirati

for a little bit of time.

duže od trenutka."

As I said, it's a very difficult process

Ponavljam, jako je težak proces

to imagine this.

ovo zamisliti.

And the great thing about it is that

I moćna stvar oko toga jest

we've only given the software one instance

da smo dali softveru samo jedan primjer

of how he thinks about "disappear."

kako on misli o "nestati".

As there is a machine learning algorithm in this --

Postoji algoritam stroja koji uči --

(Applause)

(Pljesak)

Thank you.

Hvala.

Good job. Good job.

Dobar posao. Dobar posao.

(Applause)

(Pljesak)

Thank you, Evan, you're a wonderful, wonderful

Hvala, Evane, ti si divan, divno si

example of the technology.

prikazao tehnologiju.

So, as you can see, before,

Kao što ste mogli vidjeti,

there is a leveling system built into this software

postoji višerazinski sustav ugrađen u ovaj softver

so that as Evan, or any user,

kako bi Evan, ili neki drugi korisnik,

becomes more familiar with the system,

kada se priviknu na sustav,

they can continue to add more and more detections,

mogu nastaviti dodavati sve više obrazaca

so that the system begins to differentiate

kako bi sustav mogao početi razlikovati

between different distinct thoughts.

različite obrasce misli.

And once you've trained up the detections,

I jednom kada ste uvježbali obrasce,

these thoughts can be assigned or mapped

te misli se mogu dodijeliti ili prebaciti

to any computing platform,

na bilo koju računalnu platformu,

application or device.

aplikaciju ili uređaj.

So I'd like to show you a few examples,

Zato bih vam voljela pokazati nekoliko primjera,

because there are many possible applications

jer postoji jako puno mogućih primjena

for this new interface.

ovog novog sučelja.

In games and virtual worlds, for example,

U igrama i virtualnom svijetu, na primjer,

your facial expressions

vaš izraz lica

can naturally and intuitively be used

može prirodno i intuitivno biti korišten

to control an avatar or virtual character.

za upravljanje avatarom ili virtualnim likom.

Obviously, you can experience the fantasy of magic

Očito, možete iskusiti čaroliju mašte

and control the world with your mind.

i upravljati svijetom svojim mislima.

And also, colors, lighting,

I također, boje, osvjetljenje,

sound and effects

zvuk i efekti,

can dynamically respond to your emotional state

mogu dinamično odgovarati na vaše emotivno stanje

to heighten the experience that you're having, in real time.

kako bi naglasili iskustvo koje imate, u realnom vremenu.

And moving on to some applications

Evo nekih primjera

developed by developers and researchers around the world,

koji su razvili istraživači širom svijeta,

with robots and simple machines, for example --

s robotima i jednostavnim strojevima, na primjer --

in this case, flying a toy helicopter

u ovom slučaju, letenje igračkom helikopterom

simply by thinking "lift" with your mind.

jednostavno razmišljajući o uzletanju.

The technology can also be applied

Tehnologija se može primjeniti

to real world applications --

u svakodnevnom životu --

in this example, a smart home.

u ovom primjeru, u pametnoj kući.

You know, from the user interface of the control system

Znate, od korisničkog sučelja kontrolnog sustava

to opening curtains

do otvaranja

or closing curtains.

ili zatvaranja zastora.

And of course, also to the lighting --

Naravno i osvijetljenje --

turning them on

paleći

or off.

ili gaseći ih.

And finally,

I konačno,

to real life-changing applications,

aplikacija koje mogu izmjeniti život

such as being able to control an electric wheelchair.

poput toga da smo sposobni upravljati električnim kolicima.

In this example,

U ovom primjeru,

facial expressions are mapped to the movement commands.

izrazi lica su povezani s naredbama za kretanje.

Man: Now blink right to go right.

Muškarac: Sada namigni desnim okom kako bi išao desno.

Now blink left to turn back left.

Sada namigni lijevim okom kako bi išao lijevo.

Now smile to go straight.

Sada se nasmiji za ravno.

TL: We really -- Thank you.

TL: Mi stvarno -- Hvala vam.

(Applause)

(Pljesak)

We are really only scratching the surface of what is possible today,

Mi samo grebemo po površini onoga što je danas moguće.

and with the community's input,

A s idejama zajednice,

and also with the involvement of developers

i uključenošću developera

and researchers from around the world,

i istraživača širom svijeta,

we hope that you can help us to shape

nadamo se da ćete nam pomoći oblikovati

where the technology goes from here. Thank you so much.

kamo ova tehnologije treba ići dalje. Hvala vam puno.