download keyboard samsung ori

>> david: so thank you all so much forcoming, as you've been tuning in online as well. we're so glad to be joined by ourfriends dan and michael here from leap motion, whose company leap motion hasvery generously donated some hardware for the course sot that you guys cando cool things with this device. without further ado, dan and michael. >> dan gill: all right, thanks. thanks, david. thanks folks.

nice to see you all. my name is dan gill. and as david described,i'm with leap motion. this is michael sutherland. since about 1/10 of our company isactually named michael we call him kiwi, and once you hear the accentyou'll understand why. >> but we're thrilled to be here today topresent to you folks and you folks online, and tell you a little bit moreabout leap in the development environment and our developer community,and how it may be able to

impact some of the things you'redoing as you get towards the end of the semester. i'm going to start off with just a briefintroduction for context about the company and some of the industryexamples that we're seeing and a little bit about it. and then we're going to jumpright into michael's bit. >> michael runs our developer community. so there's 70,000 plus folks that haveaccessed our developer portal and gotten access to the leap softwaredevelopment kit and are actively

building all different typesof applications that michael will talk through. personally, i run ourenterprise business. which means all the commercial usecase for the leap in a variety of industries, and i'll talkbriefly about that. >> so what i wanted to do, at a highlevel most of you are probably familiar with some form of3d motion technology. you've seen the kinect, or you've seenthe commercials for the samsung phones with the swiping.

and at some point you've probably triedone of these platforms or seen them in action. >> we really feel like we'vebroken new ground. and it's all mapped, so it's a verysoftware-driven solution. it's all proprietary algorithm based,and it's allowed us to do a number of things that are different thanthe others in the space. >> we've reached a level of accuracy thatyou'll see in the demonstrations that's far beyond what otherfolks have been able to do. so accurate to 1/100 of a millimeter, wecan track the palm position in the

fingertips for as many as fitin the field of view. and again, that accuracy level'sopened up a lot of application opportunities that haven'texisted before. it's entirely embeddable. >> because it's such a software drivensolution, this little piece of hardware that you seehere is the device. these are going to be availableto you students. we have 30 that we've donatedto the group. if you decide to do a project based onthe leap, you'll be able to take these

out on loan and spend time with them. >> we also just announcedan embedded version. so hp will actually startshipping laptops-- or has started shipping laptopsof the device, an embedded version of this in it. and because of the software and howsoftware driven this is, the hardware is incredibly simple and michael'sgoing to walk you through what's in it. we've created a very contentrich environment.

>> so those 70,000 developers arebuilding applications. you folks will have opportunities tobuild applications and possibly even get them into our applicationenvironment-- it's called airspace. we've got north of 100 apps there,and many, many, many more in the certification process and in processto being put to use. >> you'll see that it's incredibly powerfulin terms of its speed. so you'll see there's no latency inthe reaction between when you do something in the field of view and whensomething happens on the screen.

if you've played around with the kinector others, you'll notice a bit of a lag when you actually have amovement and something happens. this powerful lack of latency makesa huge difference in what you can actually do with this platform. >> and then one thing that we think isreally important, we wanted this platform to be accessibleto everyone-- as many people as possible. having such a great software drivenplatform has made the device and the platform very inexpensive.

it's a very simple piece of hardwarewith very simple, commoditized components in it that allows us to keepthis very inexpensive and very accessible to anyone who wants totake part with the platform. >> so as you'll see, there's really threemain components in the platform. we've got the peripheral that italked about, the controller. it has a usb connection to connect toany type of computing device with the supported software. we've got our software development kitin drivers, so there's a set of drivers that run on the windows or macmachine that you're connected to.

>> and then we've got the softwaredevelopment kit that we've made available to everyone at no cost onour website to access and build applications with. and then finally we've got our airspaceapp store, where there's third party applications-- both free and paid for applications. and it's everything from personalproductivity to artistic, like music and painting and drawing, to businessapplications like cad software and others.

>> so i'm going to go throughthese quick. but i thought it would be useful as youthink about the platform and think about how the languages that youcan develop in and how that applies to the platform. just to run through a couple examplesfrom the industries that we're seeing. so these are areas where people aremaking use of the leap platform to improve applications or improveoutcomes in those industries. >> education, as you can imagine,interactive displays, integrating with curriculum, like you folks are doingwith your computer science curriculum.

lots of applications forspecial needs students. so folks who can't, either for physicalor cognitive limitations, interact through a keyboard and mouseare now able to get social interactions, interactions with computerapplications, and do things they never could before. >> and then a lot of universities doinguser interface research for various industries have made greatuse of the device. health care is an exciting sector forus that i thought would be important to touch on briefly.

you can imagine sterility is of the mostimportance throughout all aspects of health care. in an operating room today, a surgeonmight have to have an extra person there, or unscrub and take their glovesoff to be able to manipulate mri images or ct scans or importantpatient information while they're in surgery. very inefficient, could compromisesterility if it's not done right. >> this type of environment now allowsyou to interact with computer applications in a completely touchlessmanner, as you'll see in the demos,

with gloves on. so they can leave thesurgical gloves on. they can access the images and doeverything they need to do in a much more efficient way. we think there's some importantapplications there. information access in what icall "germ-rich" areas-- so hospitals, atms, all different typesof areas where you don't want to touch something but you wantto access information, this has become important.

>> measuring regression asa result of a disease. if someone's losing mobility in theirhand or their arms, or in movement, being able to measure that becauseof the level of accuracy. or also progress-- so if you want tomeasure the progress of a drug or the progress and recovery from somethinglike a stroke, you can very accurately do that. and so those are some examples. >> data visualization is anotherinteresting space. i'm sure you've heard alot about big data.

everybody talks about big data. well, those large data sets in variousindustries have created real complexities around user interfaces. and how do you interact with that dataand find correlations, be able to find actionable information, be ableto share it with colleagues-- it's a huge challenge. as the amount of data grows, thatchallenge only gets bigger. >> 3d navigation with natural handmovements becomes a really interesting opportunity in that world, and we'veseen a lot of input there.

manufacturing is another one. we're going to show you some videos fromthe folks at spacex using this in the manufacturing process. also, the manufacturing floor is reallydirty, and so they've destroyed mice and keyboards, and touchscreensaren't a great solution. but they need to access things liketheir erp systems and other platforms on the floor, and it becomesa challenge for them. >> and then just a couple more beforei turn things over to michael. retail--

so hopefully at some point, you'll gointo a store and they'll have a leap enabled screen or kiosk where you canget access to a product, product options, shop online if theydon't have stuff in stock. i like to talk about it as non-intrusiveconsumer engagement. >> so i've been in sales sincei graduated from college. but we all know when we walk into aretail environment, a lot of times you get pounced on by threeor four people. we think using this type of technology,you can create real physical experiences with products andoptions and colors and different

things without having to have a bunchof people jumping on top of people when they get into the store, andcreate some interesting things. >> desktop productivity-- you'll see some basic opportunity towork with productivity apps like powerpoint to do web browsing, tointeract with your operating system. all without having you having to use amouse or keyboard, or in addition to your mouse or keyboard, being ableto get some different types of interactions. there's a lot of business applicationsthat have inefficient user interfaces,

or have interfaces that could do a lotmore if they were able to take advantage of the 3d space instead ofjust a flat 2d user interface, so we think there's a lot ofopportunity there. >> salesforce.com might be acompany you've heard of. they make customer relationshipmanagement systems. people like me in sales use themall the time, every day. but when you work with a big account,you may have hundreds and hundreds of records. and it's really hard to get a sense ofthe organizational structure, or all

the activities of what's happened insideof an account, because it's a very flat 2d user interface. so we think there's a lot of opportunityto improve the front end of various business softwares. >> and then other b2b type applications,we've seen biometric authentication. so the idea of holding your hand in thefield of view, and it's scanning your hand all the waydown to blood flow. and then being able to use that lateron for authentication to access in doing transactions, access systems, login to your laptop, control your

home automation system-- you name it, there's alot of applications. >> commanding control. so you can imagine, this is more ofthe "minority report" style thing people talk about when theythink about leap. the idea of someone in a command areawhere they've got five or six screens with video or other types of content. they need to navigate acrossapplications and call up different videos, and pull information in, anddo all sorts of interesting things.

>> and then finally cad. those environments have been a big earlyadopter of the leap platform in being able to create a more natural wayto interact with models of things that you're creating in the designprocess, or adjusting after something's been built, orthings of that nature. >> so that was just to give you a briefcontext in some of the industrial applications for the leap. i'm sure you can imagine all theconsumer applications, and if you've seen the website you knowwhat those are.

but i wanted, as you think aboutprogramming or applications that might be interesting to look into, here aresome of the areas where industries are paying a lot of attention to this. >> we were at children's hospital beforewe came here, and talking to them about a number of really interestingapplications around surgical processes and training and simulationand all different things. so there's a lot of really interestingopportunities to use the platform and to use the development environment. and so hopefully that's goodcontext for you folks.

>> kiwi's the smart one here. so i'm going to get out of the wayand let him talk you through our development community, the developmentenvironment, and all the resources that are there available to youfolks if you choose to work with the leap platform. so, thank you. >> michael sutherland: cool, thanks. so you can see there's really noshortage of opportunities there. but one of the things we see a lot ispeople sit down with leap and they're

like, where do i start? so hopefully i can go through a fewof the first steps of where to begin with all this. because a lot of people just say there'sa lot of white space, so where do i start? >> so my name's mike. as dan mentioned, i'm kindof referred to as kiwi. i'm from new zealand, as you maybe able to tell from my accent. i've lived in san franciscofor a couple of years now.

did my electrical and computerengineering degree back in new zealand, so i've sat in the sameseat that you guys are in. >> so i handle platform growth andpartnerships for our developer programs team. so i'll tell you a little bit more aboutwhat the developer programs team means in a little bit. but basically, dan mentioned, thisis the peripheral you see here. so this is the history ofwhere it came from. >> so you can see there we started backwith a very, very early prototype.

now, all that's in this, you cankind of see here a little bit. well, it's a bit hard to see on thisdisplay, but really all you've just got is a couple of infrared opticalsensors and a couple of infrared leds. the hardware is actually incrediblysimple, and that's why we're able to keep it so low cost. the magic is really what's happeningon the computer and the software in the driver layer, and that's reallywhere the breakthrough for the company came. >> so i joined and the developer programsteam started around about here,

halfway between. and what we did was these first kitsthat you see at the bottom there, they're the first developer units. and we actually sent out around 12,000of those to developers that had contacted us so that they could getstarted working with the platform. and that's really been a great seed forthe community, and we've had a lot of great stuff developedover the last year. and you'll see that when you have alook at airspace, our apps store. >> so how many of you have actuallyheard of leap motion before?

so a few of you, yeah. so that's good. >> so honestly, what was the first thingyou guys thought of when you heard of leap motion or you saw thevideos of what it does? kind of "minority report," "iron man?"yeah, we get a lot of that. >> and definitely the day will come whenwe're all sitting there commanding the world with our hands, and that'sgoing to be exciting. right now, that's not the absolutesituation that we're trying to build, but we're going to get there.

but i think it's still a goodopportunity to hear from the real world tony stark, and dantouched on that as well. >> speaker 1: right now we interact withcomputers in a very unnatural and [inaudible] way. and we're trying to create these 3dobjects using a variety of 2d tools. and it doesn't feel natural,doesn't feel normal the way you should do things. >> so we started playing around with theidea and using a few of the things

that are available out there, such asthe beat motion and siemens nx, which is what we used to design the rocket. and we wrote some codeto integrate the two. and we started off with what you seehere, which is a wire frame of a merlin rocket engine. and working through this, i can go aheadand grab it, and i can rotate it in multiple dimensions. and then what i can do is i can putanother hand in there and i can zoom in and out on the wire frame.

>> and i can also translate it. so i can move it around the screenand then zoom and translate. and this is what we startedoff with a few months ago. you can also spin itand then catch it. so this is kind of a fun way tointerface with what is really a very complex model. >> now we'll go from this to what we'reable to advance to a few weeks after the wire frame, which is to actually usea full 3d cad model of the engine. so here what you're seeingis the actual interaction

with the cad software. manipulating the real 3d modelof the merlin engine just using hand gestures. if you could just go in there anddo what you need to do, just understanding the fundamentals of howthe thing should work as opposed to figure out how to make the computer makeit work, then you can achieve a lot more in a lot shorterperiod of time. >> so then we went to a 3d projection. we started off with the kind of 3dprojection that you're familiar with

in the movies where youuse 3d glasses. we also did a free standing glassprojection, which is the sort of technology that was used inthe "iron man" movies. >> and then finally, we used the oculusrift, which is immersive virtual reality that actually tracksyour head position. and you really are movingaround the object. it feels like it's rightthere in front of you. now let's use this for an actualcomponent on the rocket, which is a cryogenic valve housing.

you can really apply your intuition andtake something from your mind to a physical object with far greaterease than we currently do. >> now that we've gotten the object out ofour head and into the computer, how do we get it out of the computerand into reality? so we're actually going to print thiswith a 3d laser metal printer. so the way that the 3d printer worksis it lays down fine particles of either titanium or [? inknell, ?] and then it goes over it with a laserand melts those tiny particles onto the prior layer.

so it builds it up justlayer by layer. so i believe we're on the verge of amajor breakthrough in design and manufacturing in being able to takethe concept of something from your mind, translate that into a 3d-- >> michael sutherland: --is that eventhough they're clearly still in pretty early stages of what they're doing withthis sort of technology, it kind of helps to demonstrate some ofthe examples that dan was talking about before. so they're really starting toinvestigate what are these next

generation uses for this sortof 3d gesture technology. so i think that's just an interestingentry to seeing how some of this technology is being used. >> so i'm going to take you througha little bit about-- it's going to be pretty high level. but we'll leave some time at the end forquestion and answers if you guys have some deeper questionsyou want to go into. but we're just talk a little bit aboutbuilding on the platform, go a little bit into the high level aspects of thesdk, have a look at what some of the

resources are that are available on ourwebsite or through the community. >> and i'll show you a few demos of some ofthe stuff that's actually out there that you can check out and give someinspiration if you wanted to use this in a project, and then how youcan approach us for help. we're here to help you guys if youwant to develop, [? sue. ?] just want to make sure thatyou're aware of that. >> so as dan mentioned,we've got airspace. so what's the real benefit for startingto build for leap motion controller?

is it just a cool piece of technology? is it a gimmick, or is theresomething more to it? >> and dan talked a bit about theindustrial applications, but on the consumer side we've actually gota really, really thriving app community as well. and you might be saying, well,another app store. so we prefer to see it asa place of discovery. >> so this sort of technology, it'sexemplified by software that is built for it.

it's not so much a system whereyou can port an existing touch application across. the biggest applications are the onesthat are built for the technology. so when you go and buy a leap motionfrom a store and you plug it in, the first thing you see is airspace. and so that's going to give you a placeto basically find all of the software that's built for platform. >> we've got over 100 apps in the storenow, so that's pretty good considering we launched just back in july.

we had over a million app downloadsin about the first three weeks. and we cover categories, productivity,games, education, creativity tools, music, science. and the store supports nativeas well as web apps. so it's a pretty good ecosystem foranyone that's buying their unit to have a lot of stuff to use. >> but for your side, on the developmentside, what that means is there's an awesome opportunity to get discovered. we shipped a couple hundredthousand pre-orders.

we're now in all the best buy storesthroughout the us, canada, uk, france, australia and new zealand, andwe're about to launch in other parts of europe. that means that everyone that buys oneof these units and gets into that store, they're going to start to seethe software that you guys are developing. so that's a pretty excitingopportunity right now. >> some of the other things that are comingup that maybe we're thinking about in a year or two.

as you go through and you developdifferent programming abilities and start to look at different types ofsoftware development, is there actually some opportunitiespost-graduation? some of our venture partners haveactually put together a $25 million venture fund called the leap fund. they've actually already funded theirfirst company, so that's been really great to see. starting to see that kind of businessecosystem building around the technology as well.

>> and shortly we're actually going to beseeing a new accelerator as well. so they'll be taking through teams, newteams that are just forming with some great ideas, and they'll beproviding them with mentors. and there's some pretty greatmentors in that program. and that will be kicking off next year,so you'll start to see some really cool stuff coming out. not just on the apps side, but interms of new businesses that are building around this technology. >> so we're providing sdk.

we're supporting both nativeand web development. i understand you guys are mainly workingin c at the moment, and you're going to be touching a littlebit into javascript in a while, so that's great. >> we've got support for c++. we do have a pure c api. it is built by the community, but ican show you how to get to that. >> so the c++, c#, objective c, pythonand java-- so if you've got any familiarity with any of those languages,there should be something

there for you to get started. the sdk's available from our developerwebsite, which i'll go through in a little bit. and then for the web development,we've got a full javascript api. >> so this is probably something that mightbe interesting as you start to go into the rest of this course. because my understanding is you'reabout to start to go into some javascript. and there's a load of great examples andtutorials on the javascript api.

so i'll walk you through some of thosethings as well, and that'll be a good platform for how to get started. >> so first is our developer portal. so this is going to be the placethat you guys want to go. if you're going to do any developmenton leap, you probably want to start here and have a check outof our developer portal. i'll just walk you through just some ofthe main things to keep an eye on. >> so this is your main download section,so this is where you're going to get your native sdk.

so that's all the languages thati talked about first-- c++, objective c, c#. inside the sdk you're going to get abunch of examples, you're going to get the documentation. so they'll be everything to get youset up for native app development. >> so basically anything that you want torun directly on your computer, that's the place for that-- not somuch for the javascript. from here, you've got accessto all the documentation. so this is going to be all thedocumentation around our api

references, as well asguides and examples. >> so you can see here we've splitit out by language. so it's pretty easy tofind your way around. and we also have, on top of this,some knowledge based articles for technical notes. this is all changing all the time. so keep an eye on it, you'llsee it evolving. >> so if we just dig into here, i'll justgive you a quick example of how we laid it out.

so if you look under here, you'vegot all the languages again. and then under c++ we'vegot our api references. so that's where you're going to wantto go to find out all the apis that are contained in our sdk. so i'll go through a few of those at ahigh level later on, but that's going to be the first stop you want to goto for getting that information. >> we've got a bunch of guides. i know it's probably difficult to seeon the screen there, but these are really just a great resourcefor you to get started.

so we've got things like how you getframe data, and i'll talk to you a little bit about what frame datameans in a little bit. all the way through to understandingthe sample applications that are-- oh, that's good. so that might make it abit easier to read. so understanding the c++sample applications-- so those are included in the sdkbundle that you download. >> so the other thing that's a greatresource for you guys, if you do start to do some development, is our forums.

you'll be able to access themup here at the top. right now these look like this. you're going to find in a week or sothat they're going to look completely different, because we're just about tolaunch a whole new forum platform. but that means an even more engagedcommunity, and it's a great place to connect with other developers that havebeen doing the same sort of work that you guys are going to be doing. >> so lots of great questions have alreadybeen answered in there. and it's a great place toask questions as well.

we're in there, our team members, ourengineering team are all in there, so great place to connect with the team. >> this is also the place that you want togo if you're going to be submitting an app, but that's probably a littlebit further down the track. but if you are interested in gettingsomething onto airspace, this is the portal to do that. you submit your app, and that willwalk you through the process. >> we have a full review team that goesthrough and reviews all the apps. there is a little bit ofa bar for quality.

we try and make sure that all the appsare really representative of what the platform can do. but at the end of the day, that justcreates a really great experience for the people that are usingthe technology. so that's kind of ourmain developer site. i just wanted to give you a quickoverview so you know where a lot of these resources are andhow to access them. >> so i also mentioned javascript api. so we've actually split out thenative and the javascript into

two separate sites. there's different ways of thinkingabout that, but we think that javascript is very unique so itreally deserves its own site. and we've had a lot of popularityfrom our javascript api. >> so this is now js.leapmotion.com. has a bit of a different look, alittle bit more fun perhaps. but this is probably one of thebest places for you guys to go to get started. >> javascript, as you'll probably startto find out, is going to be a great

language to get startedon this platform. how many of you are familiar withjavascript development already? so a couple. so what you'll find is that javascriptbeing a scripted language and not a compiled language means that you cango straight into anything that's running on the web, right-clickit, view source, you've got all the code there. so it's the easiest way to getstarted in a language. >> and what you'll find here is abunch of awesome examples.

we're adding to these all the time. you can just click on these,they'll run in the browser. >> so let's try one of them right now. so this is just runningin the browser. the code behind this is super simple. so here you go, just view page source,you get all the code, it's right here. don't worry about this toomuch at the moment. it may look intimidating, or to someof you that are familiar with it, it'll be fine.

>> but most of this is actually somethingcalled 3js in webgl. the actual part for the leap is downhere a little bit further, but it's actually very simple onceyou dig into it. and if anyone's interested, i can walkyou through some of these afterwards. but it's probably best to keepit high level at the moment. >> but anyway, so this is a greatplace to start looking at some different examples. and you can see here we've got somebasic demos to some different types of menus, a globe that you can interactwith, some data visualizations.

there's just a whole host of stuff. it's a great place to checkout some source code. >> the other thing that you'll find herewhich will be a big help to getting started is we have a greatset of tutorials. this walks you from the very basicsof just how to get a frame-- and like i said i'll go throughthat in a little bit-- through to getting a basicapplication set up. so i can definitely recommendjs.leapmotion as a great place to start if you're thinking aboutdoing some development.

and again, we've got our api docs. >> so it's a very simple layout here. it's a little bit simpler thanthe developer website. it's a little bit lighter-- it'sjust focused on javascript. but you've kind of got thosethree main things-- examples, tutorials and apis. and that will be a big help, i'dsay, for getting started. so i'll just jump back into this. so let me just grab a quick drink.

>> so this isn't really about whatis the leap motion controller. what i wanted to talk abouthere is what is it not. so a lot of people when the leap motioncame out they were like, oh my goodness, this is a mouse replacement. i never need to use my mouseand keyboard anymore. >> we don't really see it that way. because if you remember, whenthe mouse came out, the keyboard didn't disappear. the mouse augmented the keyboardexperience, and so that's really what

this technology allows you to do. it allows you to augment the experiencethat you're having. it allows you to do somethings better. >> and that's really what developingfor this platform is all about. it's not about trying to do everythingwith the leap right now. because what you'll do is you'llfind it becomes infuriating. you'll find it doesn't get theresults that you want. the best way to approach it is what arethe things that i can do better with the leap, and i'll showyou a few of those demos.

>> but what you'll start to see as you diga little bit into it, one of the favorite things for everyone to do-- and by all means, have a playaround with this as well-- is build a mouse cursor with the leap. so i want to use the leap to controlthe cursor with my finger. it's definitely an application thatcan be built with the leap. >> is it the best use of the leap? probably not. so what i would try to encourage youguys to think about is what are the

applications? if you want to build some of these,what are the applications that you can do better? they don't have to be really complex,but what are some of the things that just make your life a little bit moreefficient, or maybe a little bit more fun, or allow you to navigatea little bit better? so that's just what i wanted to giveyou a quick thought of that. >> so that leads into what are some of theapplications that are around that do a good job of demonstratingthe power of this platform?

so i'm going to take you througha couple of them right now. we can go through them pretty quickly. >> so the first one is a little game calledblock 54 which i'll put on your screen, not my screen. we'll try it windowed instead. so one of the reasons i wanted to showyou this application is because this is really something that hasnever been possible before. this is not something youcan do with a mouse. this is not something youcan do with a keyboard.

it's probably a little bit moreadvanced, but it's a great example of some of the things that youcan do with the leap. >> so what you see here, we've gota jenga tower, obviously-- or it's a block 54 tower,i should say. so what i can do here is i can actuallygrab these pieces, if my computer doesn't slow down too much. sorry, my computer seems to behaving a little bit of a hard time with this one. so i can actually grab these pieces andmove them just as though they were

physical objects. and that's really one of the majoradvantages, bringing that real-world. >> so i can literally grab thatpiece and pick it up. i can throw it away. wow, it's running a little bitslower for some reason. so you can kind of getthe feeling there. so this is something that you literallycould not do on a-- i'm just going to push thisone out of the way now. there we go.

>> so that's a lot of fun. if my computer wasn't chuggingalong so much, then that would be a lot smoother. but you can kind of see there thatthis is an example of bringing something that was real-world into thedigital space, and it's allowing you to interact in a way that'svery natural. >> i'm not using a menu systemto go through that. i'm not clicking, or using keyboardshortcuts or anything. it's just literally me reaching outand manipulating the blocks in the

digital space. >> so this next one is a little bit alongthe same lines, but it's about bringing these real-world experiencesinto the digital world. and so this was an experience that ibelieve one of our co-founders had. i don't want to quote him on this, butthere was the ability to be in the ocean and see these schools of fishswimming around and being able to interact with them. and that's something that's reallydifficult to communicate verbally. it's also very difficult to communicatedigitally without a proper

input mechanism to be able tomanipulate that 3d world. >> so i'll just bring this one up. you might not be able to see that on thestreaming version-- it could be a little dark. so what you're seeing here ismy hands in the 3d space. i've got complete freedom of movement. and i can just hold my hands still,see the fish and then scatter them away. >> and you can see the freedom in thisdigital space is like something that

really hasn't been possible before. i can bring them out to the screenand scare them away. so it's a simple demo, but it'shighlighting the fact of being able to bring some of these real-worldexperience that haven't really translated into the digital space intothis digital space finally for the first time. >> audience: you can actually see those atthe museum of science [inaudible] >> michael sutherland: and thisis a great one as well. because what we find is when peoplefirst put their hands into this, it's

the best time that they've seenthemselves represented so fluidly in their digital space, so you usuallyget an interesting reaction. so by all means, if we've got timeafterwards, i'd be happy to show you guys some of these demos. >> so this next one-- i won't bring the slide deck up again. this next one is about creating anexperience that you can just explore. so there's very limited rules to this. it's creating this immersiveexperience.

>> and the developer that built this is aguy called eddie lee out of japan. and this was actually an experiencehe had in kyoto. and he wanted to basically bringthat experience and share it with other people. >> i don't know if you can hear that. but you can just drag your fingersthrough the water and mess around with the reflections. there's nothing that's tellingme what i can do. it's just a very zen experience.

but it's something that you can justreally immerse yourself into and forget about how you're actuallyinteracting with it. and just put your hands in andjust feel your way around. >> and you can see the entire environmentis there to just play around with. and there is actually astory line to this. it takes a little while to go throughit, but you can explore your way through it. it's kind of a lot of fun, anda lot of natural interaction. >> so this is actually somethingby the same developer.

this one's a little bit crazier, butit highlights some interesting use cases, and again, something thatyou can only really do with this sort of platform. so i'll just come around here. >> so this is actually the menu page. and this is a menu like hasn'treally existed before. so literally it's just looking at howmany fingers i hold up, and basically choosing the menu through that. so you can see the rules arebeing rewritten around

interface design here. you've got total freedom todo whatever you want. >> so in this one, this is kindof a little crazy. i love these guys. i could watch these guysbouncing all day. he's got a whole bunch of differentexperiences there. >> all of these you can getthrough airspace. so feel free to sign up. it's free.

you can go have a look at the apps. >> this is kind of a musical experiment. but what he's doing is he's using thefull 3d space to create different sound effects. it's probably a little bit hard tohear through the sound system. but basically he's using this full3d space to create a new type of instrument. and then whether i use three fingers,four fingers, or one finger, i can basically start to changethe effect of a sound.

so it's definitely very experimental,but it highlights that freedom in that 3d space. >> so you saw block 54, it's a game. and the last two were more creative,experiential kind of things. it's easy to get wrapped up in that kindof creative world, and there's so much amazing stuff thatwe're seeing come out. and if nothing else, that's a greatreason to develop for the leap motion. >> we're seeing so much amazing, creativestuff, but there's also an element of efficiency.

and so i just want to show youa quick integration that the google earth team-- some of you might haveseen this before. hopefully i've got an ok connectionhere, because it is a little bit bandwidth intensive. >> but you're probably familiar with googleearth and how you generally would navigate around that. it's click and drag, you've got thesliders for zoom in, zoom out. if you're really proficient at it,you've got keyboard shortcuts, click

and pan and tilt. there's all these different ways thatyou can navigate around this 3d environment. but what the google team did was theyjust rewrote the rules on that. >> we might have a little bitof bandwidth issues. but what you can kind of see there isyou can basically just navigate. so let's see where we want to go to. oh, yeah, we're a littlebit stilted there. but what it's allowing me to do isi can control multiple degrees of

freedom all with one fluid movement. >> so i can pan left to right. i've got look up, look down. i can change my elevation. i can move forward. i can basically go and i can spin arounda certain point and just keep my focus on it. i've got complete freedomin this 3d space. >> and all of a sudden, my efficiency ofnavigating around the space is just

multiplied immensely. so i can jump from boston to sanfrancisco to new zealand in a couple of seconds. previously, that sort of operationwould have taken me quite a few different clicks and movements andkeyboard shortcuts, and i have to remember it all. >> so this is an application where thissort of natural interaction is allowing a greater efficiency. so that's another thing to keepat the back of your mind.

is this something that i can make moreefficient in what i'm building? >> and the final one i want to show youbefore i go into the sdk is about the educational possibilities. and this is something that dantouched on lightly with the children's hospital. we're going full screen. hang on a second. here we go. >> this is a bit strange.

let's try giving thata full screen again. well, that one doesn't look like itwants to run on this predictive for some reason. interesting. oh well, that's too bad. >> so this one is basically a little app. you can have a look at it later,after this if you want. but basically what it's doing is it's afull 3d representation of the skull. and what you can do is basicallytake it apart in 3d.

it becomes a 3d jigsaw. >> so some of these applications, they'll[? ask ?] for a more immersive learning environment. so when you're able to interact withwhat you're learning, you start to take it in a lot more. so we're seeing a lot of interestingapplications being developed, both in early learning and special needslearning, as well as all the way through the sciences. so there's a lot of interestingapplications along that.

>> i'll try and show you this one, but it'sgoing to be a little bit hit and miss whether we can getit to work here. this is only showing half ofthe anatomy at the moment. but what this is allowing you todo is see how you can navigate around this in 3d. i can basically start removing sectionsand be able to navigate. i can actually start to basicallypeel back the different layers. it's almost like seeingan mri in real time. >> this is part of the biodigitalhuman project.

so this is actually something that'sbrand new that's come out. you can sort of see how you can justtake pieces apart and then just basically navigate in and examineit a lot more closely. you can see this is actually runningdirectly in the browser, so this is an example of what's possiblewith the javascript api. >> so those are a few different examplesof some of the applications. you saw the creative exploratorysituation. you saw some of the efficiencyincreases, some sort of interactive learning examples.

so you can see there's a widevariety of different applications, different software. >> i'm guessing that probably not manypeople have had a look at the sdk by this stage. so i'll just go at a very, very highlevel through what is the data that's making all of this work. what is it as you as a developer wouldbe working with to create those sort of experiences. >> so i've touched on, a fewtimes now, frames.

so at the very, very lowest level,we have what we call frames. and a frame is basically returned to youup to 200 times a second, and it contains everything thatthe leap sees. so the leap see hands, it seefingers, and it sees tools. so i'll show you quickly in ourvisualizer what that looks like. >> so if you're interested in doing someleap development, this tool here is probably going to be one of the mostuseful things that you can start to play with. it's actually not immediately clearpossibly where you get to this from.

and i'll show you quickly just so thatyou're all aware of where you can actually get to this. >> so when you're running the leap motionsoftware, you've got this little icon up here. this is where you can launch airspacefrom, it's where you can get to your settings from. one of the things in here is thisthing called the diagnostics visualizer, and that's undertroubleshooting. it'll launch this tool here,and this is basically--

>> audience: [inaudible]. >> michael sutherland: oh, right. yeah, thanks for that. i noticed that as well. thanks. it just kind of popped out before. thanks for noticing that. >> so this is basically just whatthe leap is pumping out. so this is the data that's comingout being processed by us.

and at the end of the day,this is what you get. so this is hands andfingers, basically. what you're seeing there is all myfingers represented in real time down to a hundredth of a millimeter. >> you can see the arrows. the arrows represent thedirection of my finger. so that's something thatyou'll get through ipi. you can see where they're drawing--that's the position of the fingers. and you also get the velocityat any one time as well.

>> and you'll see there the two bigcircles representing my palm. and you've got a big arrow sticking outthe bottom, or the top if i hold my hand upside down, and those arerepresenting the palm normal vectors. so basically normal vector being just avector that's sticking straight out of your palm's surface. so those are the fundamental buildingblocks that you'd be working with with building leap software. and this tool allows you to reallysee exactly what's going on. >> and there's a few things that you cando, a few little tips in this

visualize that may be helpful. one is just to represent your fingersa little bit more clearly. the other thing that may helpis this is essentially what the leap is seeing. >> so i mentioned before, there'sa couple of optical sensors. so these things have basically a fieldof view of about 150 degrees. and so this yellow box here isrepresenting what this can see. so you can see here as igo outside that box i'm starting to lose my hands.

and if i start to go outside here, it'llstill pick it up, but you're starting to lose it on the edges. so this gives you a bit of a senseof the space that you've having to play with. >> and you'll see here if you pressh, it'll toggle this menu. and that will actually give you a wholebunch of different options that you can access. most of it you probably won't need. but it's a great way of visualizingwhat's going on without getting dug

down into the data. >> so that's the frames, hands,fingers and tools. actually, i'll show you just quicklybefore i go back, the tool. so let me see, this should work. so you can see here my hands. and if i bring in this pin,it's coming up as gray. and what that's saying isthat that's a tool. >> so we have what we call a tool api. it actually recognizes objects like pensor paintbrushes, anything that's

roughly of this sort of shape. it'll be able to tell that that's nota finger, and you can actually use that to your advantage when you'rebuilding software for this. >> you can start to bring real-worldobjects into your application, for example. so you might have literally an easelof different paintbrushes, and each paintbrush you've coded upto represent a different brush in the software. so instead of actually changing itthrough a menu on the software, you

can literally just pick up a differentbrush and start painting with it and have the software adapt to it. >> so that's frames, hands,fingers and tools. so that's the really, really lowlevel stuff that we have. i don't know if anyone here has anyfamiliarity with the kinect. one of the things we get asked alot is where is the raw data. and what that means is in other 3dtracking systems, it's basically a blob of data. >> what we do here is create amore structured approach.

so this is actually thelowest level data. and we find that because it's structuredlike this, it really helps people to get started quicker. if you just got given a full 3dblob of data, it becomes very difficult to work with. so that's one of the reasons why it'sstructured in the way it is. >> so is there any questionsaround any of that? nope. so we'll move on.

>> you might be able to start to see eventhough the data is structured in a way that gives you literally what you'reseeing, like hands and fingers, it can be a little daunting at first to startto figure out how to work with that. now i'm tracking fingers in 3d,what do i do with that? so we do have some higher level apisto help get around some of those areas, and it might be a way to getstarted a bit quicker as well. >> so this is conceptuallycalled motions. it's a part of our api. you'll find some guideson what motions is.

but at a conceptual level, what it'sdoing is it's basically taking all these movements in the space andturning them into one of three things-- translation, rotation,and scaling. >> so don't get too caught up in that. but what it basically allows you to dois it converts these complex movements into single digits, or a degreeof rotation, or a scaling factor like a number. so what it does is it abstracts out alot of the complex mechanics and if i'm doing this, it gives you a numberthat says this is scaling by 10.

>> so what you can do then is if you hadan image that you wanted to enlarge, you could grab the image. and then use this sort of api to say,well, now i'm scaling it by a factor of 10 and you don't have to worry aboutall the data that's going on. so it's something to just keepat the back of your mind. it may make it a little easier if you'retrying to do some of those more complex interactions. >> that's what we call the motions api. you won't see it actually calledmotions in the documentation.

it's actually a collection ofapis from different places. but what i can do is if anyone'sinterested in learning more about that, i'm happy to point you at someguides for how to get started on that. >> and then the next thing up, which isthe thing that most people are probably most familiarwith, is gestures. so this is much higherlevel abstraction. so you're basically taking all of thesesort of movements and you're saying, right, what's a discretething that i can do? so i can circle with my finger, or i canswipe with my hand, or i can tap

in the ear. and so we've broken those down just totry and make it a little bit easier to get started into someof these gestures. and i'll show you show a little bitabout how some of those work. >> so back in the visualizer,if i turn gestures on-- let me just stop that so it's notrotating and making it run dizzy. right, there we go. turn it up. right, here we go.

>> so now you can see that myhands are in the space. if i draw a circle, it's comingup and showing a circle. and you can see that's actuallyin any plane. it doesn't really matterhow i draw it. but it's basically detecting that i'mdrawing a circle with my finger. >> and at the api level, we've tried tomake that as easy as possible to use. so you don't really have to think aboutthe mechanics of tracking points in 3d and figuring out whetherit's a circle. you can just say, isa circle happening?

so it's one thing that you canuse as a control mechanism. >> you'll also see there thatwe've got swipes, taps. you can see those little ballsbouncing there at the bottom. so those are visually how weshow what the gestures are. >> but in terms of developing software forthis, it's just a high level way of simplifying a lot of the complexityof tracking fingers, so you can just use those as is. you'll start to see differentapproaches to that. and i'll show you another approach alittle bit later on of a different way

of doing that, but thoseall built into the api. >> so we have a few other partsto the sdk that might be interesting to you guys. so just to cover those last bits,there's three levels of abstraction i talked about. the low level, which is the frames,the hands and the fingers. >> the middle level, where it's convertinga lot of that movement into continuous movement, so rotation,or scaling, or translation. and then to the next level upwhich is the gestures, like

am i doing a circle? am i doing a tap? am i doing a swipe? >> then on the other side of it, we've gotthings called the interaction box. i don't want to go into too manydetails, because this is all just to give you a bit of a taste of whatsome of the things are. you're definitely more than welcome toreach out to me with specifics later on about this. but the interaction box is another waythat we're trying to make it a little

bit simpler to think aboutcoordinates in the space. >> so i mentioned beforeyou saw the space-- it's this 3d inverted pyramid. that could become a little tricky. you can sort of see it there. that can become a little tricky ifyou're trying to translate that into screen space where you're displayingwhat you're working on. >> so what we created isan interaction box. it's going to be very difficult--

oh, there we go if i turn that one on. so you can see that white box there. and basically what that's doing is it'sjust mapping that to zero to one, zero to one. so you just get a scaled spacethat's always fixed. you don't need to worry about howfar you are above the device, or wherever it is. >> and this adjusts-- at least it should adjust.

this is obviously demo mode. but basically what will happen withthat is it will just adjust to wherever the person is above the device,and it'll create a consistent space for you to work in. >> it sounds a little complex with theway i'm explaining it there. but what it essentially allows youto do is just forget about where the person is. it just gives you a scaled zero to onein the y, zero to one in the x. and you just don't need to worry about allthe complexity of where the person is,

whether they're using big movements orsmall movements, and it just scales everything for you. so that's just somethingto keep an eye out. >> if you do look through the documentationand you see something about interaction box, that'swhat that's referring to. it can be a little difficult conceptto understand what it is. and it's unfortunate that it's notscaling up with me, but that's ok. >> and while we're on this view, the otherapi that would be interesting to maybe talk about is ourtouch zone api.

so one of the first things that peopleask is how do you click with the leap? it's kind of an interesting question,because you don't really need to click with the leap. what we try and encourage is to thinkabout actually interacting with the space, grab it and move it-- youdon't need to click and drag. >> but for the applications where some sortof interaction is necessary, we have an api that's calleda touch zone api. and it just tries to take a lot of thecomplexity of figuring out exactly what the user's doing in the air andsimplifies it into just an event that

says you've either clicked or not. and i'll just show you veryquickly how that works. >> so you can see here my finger's beingrepresented as a cursor, and you'll get given this positionthroughout api. and as i start to move forward, itbasically says i'm now clicking and i can drag this around. and it doesn't really matter wherei am in the 3d space-- it'll work no matter where i am. >> and so at face value itlooks very simple.

there's actually a lot of complexmechanics around that. so that's why we try to encapsulateall that into an api and make it a little simpler for you guys. so if you have a look at the api, it'sactually pretty straightforward to build that into your application, andyou don't have to worry about all the complexities of where theperson's hand is. >> so there's a lot of otherstuff in the sdk. if you want to have a bit of explore,you'll start to find some of the other things.

but those are some of the high levelconcepts that are in our sdk. >> it might be a little bit much to take inwithout having had a chance to play around with the leap yet. but i just wanted to give you a bit ofa flavor of what's in there so that when you do get to it, as i said,feel free to reach out to us. i can point you in the right directionfor any sort of documentation to help you get started. >> so as i mentioned, our sdk hasa bunch of native languages. we have the javascript api.

one of the easiest ways to get startedmight be to look at some of the platforms of the frameworksthat are out there. >> i don't know if people arefamiliar with unity. famo.us, goo and vuo are allvery new ones to the scene. unreal you've probablyseen in game engines. >> but what these kind of environments dois it may be an easier way to help to get started. because what some of them do willprovide you with a 3d framework to start with.

so it kind of takes out someof the complexity. you get more of a visualenvironment to work in. >> famo.us is a new platform that's comingout for a web app development. their aim is to make web appdevelopment super easy. so that will be coming out-- there's no time frame forit at the moment. but if that's something you'reinterested in doing, it could be one to keep an eye on. >> goo is an amazing html5gaming platform.

they're doing a very visual editoras well online, again, for high performance web apps. vuo is something i cango over very briefly. this brings in the conceptof rapid prototyping. >> and i don't want to gotoo deep into that. but one of the things if you are reallyinterested in getting into this sort of development, finding a good toolto do rapid prototyping could be really valuable. and what i mean by that is it's aframework where you have to do very,

very little effort toget a lot of return. so you don't really needto do a lot of coding. a lot of it's very visual-- it'sdragging blocks around. >> in fact, i can show you a very,very brief example of this. so you can see here, it's just acompletely visual environment-- you don't even need to code. oh, great-- we won't do that demo at the moment. it doesn't seem to wantto run at the moment.

so without going into too many details,it's what's called a visual programming language. it allows you to get some basicfunctionality working. >> i'll just quickly bring up acompleted version of this. so you can see here, this is a verysimple application that basically takes an image and allows you tomove it around with the leap. and these green blocks here areessentially all you need to do to get started with the leap side of it. so it's a good way to get started.

if you have some ideas you want toexperiment with before you even get any code down, it's a goodway to get started. >> audience: if we use [inaudible] look at, would it translateinto actual code? >> michael sutherland: in that situation,i don't think you have the ability to translate to code. there are definitely some otherframeworks out there. quartz composer is actuallyan apple tool. it's no longer officially supported, butthere's a big community around it.

>> we've seen some amazingly leapstuff come out of that. there's some plug-ins available. i think there is access to low levelcode from quartz, although i'm not entirely sure about that. but that's a good question. >> so i'm just going to show you some very,very high level terms for things to think about when you're developing. lighting conditions generally aren'ta big issue for the leap anymore. we've got an amazing team that'sbasically been able to eliminate for

almost all lighting conditions, becausethat's something that can potentially affect. infrared sources coming in from theoutside have the ability to affect the performance. >> in general, you won't really comeacross many stumbling blocks. if in your development you see thatthe device goes into robust mode, really all it means is it may havedetected that there's some infrared light sources in the environmentand it's compensating for it. so don't be too worried about that.

>> in general, when you're designingsoftware for the leap, it's important to realize that this could be the firsttime that your user is using this technology. and this is something that it's hardto get your head around initially. what we try and encourage people todo is think about instead of just allowing the user to have to find theirway around the interaction, what you're asking your user to do, try andexplain it to them a little bit. treat them as though they may neverhave seen this technology. sometimes people won't even know toreach their hands out over the device,

so don't take anything for granted. >> if you have a look around on airspace,you'll notice that a lot of the apps really infer the user into how toactually interact with that app. that's something to just be aware of. if you are developing softwarefor this platform, it is new. people aren't familiar with thetechnology yet, and so you may need to help ease them into whateverit is that you're building. >> data is your friend. i mentioned before the visualizer.

it could be one of thebest tools you use. it just allows you to lookand what you're doing. think about the action that you'retrying to code up, and then look at what it looks like in the visualizer. and then it will give you a better senseof what that data means that you're getting out of the sdk. >> if you're doing anything that needs amenu, menus are something that you want the user to be able todo without even thinking. it's not really part ofyour application.

it's a part of how the useruses your application. >> so we have some resourceson the developer site. just a couple of different systems formenus that take the burden off you guys for having to think abouthow to build menus. because menus can be something that youcould spend a lot of time trying to build into your application, whenreally what you're trying to do is build the idea that youhave, not the menu. so i would recommend if you have to doany sort of menu systems, definitely have a look at the resources wehave on the developer site.

we've got some great examples of howto do menus, and how to keep them consistent so that users have aconsistent experience across applications. >> visual feedback. so what i mean by that is if for exampleyou're trying to do something that is showing a 3d space, it's veryimportant to provide some sort of visual feedback. so whether that's showing where yourfingers are in that space, or in the case of block 54 that we saw at the verystart, you may have noticed that

the [? pedals ?] were illuminated. and when i went close to the towerof blocks, you could actually see visually that i was close to them. and it's a small trick, but it'sactually a very important one. so make sure that the user'soriented in that 3d space. >> and again, rapid prototyping. if you can find some tools that youfind helpful, i definitely would encourage you to investthe time in it. being able to get your ideas out quicklyinstead of having to spend a

lot of time coding at a lower level andtrying to figure out how to code it up, if you can get those ideas out infront of you, play around with them a bit and then code it up, itcould be a great time saver. >> so we're getting to the end. how are we for time? >> speaker 2: [inaudible] eightminutes until 5:30. >> michael sutherland: we'llfinish at 5:30? >> speaker 2: i do that. that was the spot weadvertise in here.

but we can do one onone q&a after this. >> michael sutherland: yeah, i won't go toodeep into the rest of this then. i did mention before a differentway of doing gestures. if you're interested and you are workingwith javascript, this is a javascript application that a developernamed robert leary built. what it does is it takes a lotof the complexity out of recording and using movements. >> so what he's done is he's basicallycreated a gesture recorder. you can type in the gesture,record it.

it spits out something that you canthen pull into your application. so instead of having to code up all thecomplex movements in 3d, you can just take this, do the action, andsave it for your application. so that could be an interestingtool to help you get started. >> i can go through these very quickly. it's just a couple of videos that showsome of the interesting applications. some of the stuff you might not seeeither in airspace, but it's floating around in the developer community. just some amazing work that people havebeen working on that show some

maybe some more unique applicationspossible. >> so this is a system that's usingbasically head tracking on the camera to give that depth perspective. but you can see it's a prettyinteresting visual trick. so that's kind of an interesting thingthat is possible with this type of technology. >> and then some of you may be familiarwith the oculus rift. this is just some experimental work thatsome of the developers have been doing around combining leap motion withthe oculus rift, so for the first

time you can be insidethat virtual world. so that's going to be an interestingapproach for gaming coming up soon. the oculus rift is a vr headset. poor guy-- he really got a hard time. >> this was an exhibit that was doneusing projection mapping with the leap motion. just a really nice interactiveenvironment where people can just play around.

you can see there they created these3d trees using projection mapping techniques. this was an interesting one done intaipei with heineken doing an installation. >> the whole "iron man" approach there. but for the sake of time, i'lljust quickly get to the end. so if there's three things that wouldbe nice to take away-- because i realize there's a lot of informationthat we just covered. and a lot of it you'll really need tospend a little bit of time to just dig

down into the resourcesthat are available. >> but i think the first thing is really ifyou are designing software, try and design for the user, not for "iron man."so forget about i want to be "iron man." i want to be on "minorityreport." but instead, design for the user. design for the person that's goingto be using your software. >> so think about how can i maketheir experience better? how can i make something thatthey're doing better? and that's really going to be wherethe most powerful and the most

engaging software comes from. >> and if you're familiar with ui/ux, youcan almost throw the rules out the window in some sense. with this sort of technology, we'restarting to rewrite the rules as we go, and that just means thatyou've got a blank canvas. so you guys are really startingat the right time. if you're just getting into programmingnow, that means you get to write the rule book as you learn,so that's an amazing opportunity for this.

>> and i would just say again, be able tofind a way to prototype quickly and then build. don't necessarily waste all your timegetting into the nuts and bolts straight away. see if you can get your ideas out. >> it used to be that it was goodto get them on paper. and paper's still a great way to go. but once you start to get these dynamicinterfaces, you really start to need some better tools to be ableto get those dynamic ideas out.

and so if you can find some tools thathelp you to prototype, try and learn them and use them, and you'llprobably save yourself a lot of time and hassle. >> so a few resources. once you start gettinginto javascript, js.leapmotion.com/tutorials, that'llbe a great way to get started. examples again on js.leapmotion.com,you'll find some great javascript examples. >> please feel free to engage in theforums, ask developers, ask us.

it's a great way to learn. if you're interested in reading morecontent about what's out there, some of the thought leadership in the space,labs.leapmotion.com's a great blog for that. we're putting out new content everyweek, it's a great space. >> and if you want to connect withus, again, the forums. you can email us atdevelopers@leapmotion.com. we're on @leapmotiondev on twitter,so just tweet at us. we're pretty active on there.

and our main handle, @leapmotion. @leapmotiondev obviously is ourdeveloper twitter handle. >> so that's really about it. if there's some time for questions,definitely happy to answer any questions. if you think of anything afterwards,please feel free to reach out to me directly at kiwi@leapmotion.com,or tweet at me at @kiwi. cool. any questions?

>> audience: in addition to developing appsthat [inaudible], how feasible is it to make [inaudible] level software sothat you could scroll left, right, up, down, and any [inaudible] applications, [? for instance ?] [inaudible]? >> michael sutherland: so thereare applications for that. if you have a look on airspace, you'llfind a few different applications. some of the more popular ones are onecalled handwave that allows you to do some basic gesturing.

if you want to do that sort of stuff,there's really nothing that limits you in the sdk to do it. it's really a question of if you'rebuilding that os level control, is it actually making that experienceof using the os better? >> over time, we'll start to see theoperating systems evolve to a state that really is made forthis type of input. for right now, we are actually usingoperating systems that have been built for 26-year-old technology. if you have a look at the mac interface,it really hasn't changed in

about 26 years. so we're really battling a 26 yearlearning curve where people have got so used to this type of interface thatit's hard to see beyond that. so if you can improve that experience,that's a definite win. >> but if it's just doing a gesture forthe sake of doing a gesture, what you'll probably find is that users,they'll find it easier to just go back to their keyboard and mouse becausethat's what they're comfortable with. so that's why it's really important tothink about who am i designing for? who is that end user, and how cani make their life a bit better?

>> but if we have the time, ican show you a quick one. this is an interesting onethat just came out. it's a very simple cursor, but it'skind of cute because it has this little hand-- or it doesn't. are we in there? >> well, i won't show you that demo. but that [? leapcursor.js ?] is an interesting little example thatbasically lets you scroll up and down

fluidly in a web page and sort of clicksjust by flexing your hand. so it's supposed to be more ofa laid back kind of scroll up and down, flicks. >> audience: you mentioned the deviceitself is largely commodity hardware. what is the underlying hardwaretechnology that's actually doing the detecting of objects? >> michael sutherland: so the actualdetection of the objects-- so basically if you were to hack the usbon here, you're just going to get a whole bunch of image data back.

people have already done it. where the magic is happening isbasically once it gets into the computer, it's essentially someproprietary algorithms that were originally developed by our co-founderand have now just taken on a life of their own. >> audience: is it through infrared,or a magnetic [inaudible]? >> michael sutherland: so it'sjust purely infrared. so literally, it's kind of like havinga little webcam sitting on your disk and then a spotlight shiningon your hand.

it's just all done in infrared. so it's just some infrared opticalsensors, and some infrared leds, and there's really nothingtoo complex about it. it's the way that we're able to takethat data and then turn it into something useful in 3d. >> audience: so it seems fairly easyfor people [inaudible]. but is there any way for a developerto maybe apply [inaudible] for other types of objects-- maybe facesor other types of things that the user might put forward?

>> michael sutherland: at the moment wedo support a limited set of tools. unfortunately, with the way that we'vestructured the data, because we wanted to do it in a simple fashion or one thatmakes the most sense for hands and fingers, the api won'tsupport face tracking or generic object tracking. that may come in the future. but for right now, it's really finetuned for hands and fingers and specific tools. >> [? dave: thank ?] you so much.

this is terrific. >> [applause]

Share this post