This post violates a couple principles I like to keep in mind when blogging. First, it doesn’t have an image, and second, it doesn’t split the main body from the intro, offering a “read more” link.

But, it also has something previous posts on TRW do not have: the the distinction of having been edited and finalized in Post’er, the native WordPress app for webOS and my Palm Pre.

Post’er website: http://poster.nizzoli.net/

Local Bytes

• The Nutter administration announced that the City is buying that municipal wifi network that has been a non-starter from the beginning. They will use it for surveillance and traffic cameras, enabling employees to file reports from the field, and other City purposes. They will invest about $17 million in the next 3 years to improve the network. (Source)
  • Engadget mentioned this story this morning, too. (Source)
• Technically Philly has their always-informative Comcast Roundup. (Source)
• Philadelphia Tech News mentioned the currently-alpha Appify, reported on originally by the New York Times. Basically, Appify is a directory of city-specific applications for Apple, Palm, Android, and Blackberry devices. No access yet to the directory itself, but developers can add their app. Hopefully there will be a strong showing for Philly. (Source)

Everything Else

• DailyTech reports that the Playstation 3 will support the newly-announced Blueray 3D specification. (Source)
• Ars Technica mentioned that the next iteration of Ubuntu Linux will include a panel overhaul and a new social networking component. Always exciting news for nerds like me, who use Ubuntu as their primary OS.  (Source)
• Engadget posted about how Apple stole Johnny Chung Lee’s head-tracking idea, which I first saw him present on TED. (Source)
• The New York Times Bits technology blog mentioned that the Google phone’s name, Nexus One, sounds like a reference to the androids in Phillip K. Dick’s story Do Android Dream of Electric Sheep?, on which the kickass 1982 flick Blade Runner was based. (Source)

Edited at 5:30pm, December 17, 2009 to add categories, links and tags.

[Johhny chung Lee], eat your heart out. Check out what these guys are doing with face tracking and immersive 3d as their final project in class. They’re using a singe camera and an FPGA to produce the demo you see in the video. Facial tracking is done by skin color, so that might have some issues in some environments, but being able to have perspective shift with you, without rigging up some more hardware is fantastic.

We realize that this is completely different that what [Johnny] is doing. We love [Johnny]’s work and think it is ground breaking to be able to pull this stuff off with a cheap game controller. We just couldn’t help but draw the parallel from the end result.

[thanks Bruce]

An entirely controller-free gaming experience for Microsoft’s Xbox 360 will be commercially available as early as Christmas 2010 for as cheap as $80, according to recently leaked information from a secret Microsoft tour in the United Kingdom. Known as Project Natal, the new technology will allow gamers to seamlessly interact with virtual worlds and characters by speaking, gesturing and presenting physical objects to a 3D camera.

According to the inside source, Microsoft already has 14 games in the works. One is called Ricochet, in which a player uses his entire body to bounce balls at blocks and score points. In Paint Party, gamers create an onscreen masterpiece by posing in the shape of a desired stencil, choosing colors through speech recognition, and splashing paint onto a virtual canvas with throwing motions.

This project vision video features some other possibilities, including martial arts between a gamer and an onscreen character—without the need for cumbersome sensors—and the ability to embody a Godzilla avatar on a destructive rampage.

In this earlier video, a woman interacts with a realistic virtual world and a pixilated boy named Milo, who can recognize the emotion in her voice, hold a conversation and even comment on a drawing she holds up to the camera.

The wizard behind these unprecedented gaming features is a small black box called the Project Natal sensor, which is compatible with existing Xbox 360 consoles. The magic box relies heavily on sophisticated software for human-computer interaction. “The human tracking algorithms that the teams have developed are well ahead of the state of the art in computer vision in this domain,” Johnny Chung Lee, a researcher in the Applied Sciences group at Microsoft, wrote on his blog. “The sophistication and performance of the algorithms rival or exceed anything that I’ve seen in academic research, never mind a consumer product.”

Working together, a depth camera, a multi-array microphone, and a set of specialized sensors enable full body 3D motion capture, as well as face, voice and object recognition. “Essentially we do a 3D body scan of you. We graph 48 joints in your body and then those 48 joints are tracked in real-time, at 30 frames per second,” project director Alex Kipman said in one interview. Not only will Natal register just how forcefully you kick a virtual soccer ball or how skillfully you jump on your digitized skateboard, it will recognize how many people are playing with you, distinguishing between familiar gamers and guests. Depending on how far away you are from Natal, the sensors can even differentiate the individual fingers on your hand.

The applications of such interactive technology are not limited to videogames. “These are not just the kind of experiences that you’re going to see with the 360,” Phil Spencer, general manager of Microsoft Game Studios, told The Seattle Times. “It’s a foundational event for the company, to think about how we will evolve our platforms.” If things keep going well, perhaps Microsoft will liberate laptops of keyboards and mice in time for Christmas 2011.

Signore e signori, il futuro si avvicina… chi di voi non ha mai sognato una cosa del genere?

Quello che avete visto è frutto della combinazione (aka COMBO) di Compiz Fusion e di una semplice webcam compatibile con il progetto OpenCV, condita con un po’ di pene dell’inferno per compilare OpenCV da terminale.

Premessa

Questa guida, il software utilizzato ed i plugin che andrete ad installare sono sperimentali ed instabili. Quello che vi consiglio e di non proseguire nel caso non sappiate quello che state facendo per evitare così di ritrovarvi un sistema compromesso… io ve l’ho detto!

Preliminari

Controllate se la vostra webcam è presente nel seguente elenco: compatibilità webcam. Se è presente potete continuare, altrimenti potreste comunque proseguire e controllare se il plugin funziona comunque con la vostra cam. (e ovviamente segnalarla agli sviluppatori di OpenCV!)

Iniziamo a vedere cosa serve per avere questa meraviglia funzionante:

sudo apt-get install build-essential subversion git-core libtool compiz-dev libcompizconfig0-dev libgtk2.0-dev libavutil-dev libavcodec-dev libavformat-dev libjpeg62-dev libtiff4-dev libavutil49 compiz-fusion-bcop

Download OpenCV

Procuratevi l’ultima versione di OpenCV:

svn co https://opencvlibrary.svn.sourceforge.net/svnroot/opencvlibrary/tags/latest_tested_snapshot

Dopo poco dovreste aver scaricato l’ultimo snapshot disponibile.

Installazione OpenCV

Come dalla migliore tradizione:

./configure
make
sudo make install
sudo ldconfig -v
export LD_LIBRARY_PATH=/usr/local/lib:$LD_LIBRARY_PATH

Se fino a qui non avete avuto problemi di sorta, dovreste aver installato OpenCV correttamente. Passiamo al plugin per Compiz!

Download Plugin

Per scaricare il plugin utilizzate git:

git clone git://anongit.compiz-fusion.org/users/klange/headtracking

Installazione Plugin

Al solito, siamo tradizionali:

cd headtracking
make
sudo make install

se compilando il plugin incappate in questo errore:

facedetect.c: In function ‘headtrackThread’:
facedetect.c:264: error: too few arguments to function ‘cvRetrieveFrame’

Scaricate e installate il plugin presente qui.

Risultato ed utilizzo

Ora se aprire il compizconfigsetting-manager (gestore configurazione di Compiz) dovreste avere a disposizione un nuovo plugin nella categoria extras chiamato headtracking:

E se siete fortunati.. mooolto fortunati, una volta attivato dovreste vedere le finestre spostarsi a sinistra/destra a seconda dei vostri movimenti per simulare l’effetto 3D. Buona fortuna!

Fonti:
http://opencv.willowgarage.com/wiki/

http://www.ilportalinux.it/content/installare-il-plugin-headtracking-di-compiz-su-linux#

http://forum.compiz-fusion.org/showthread.php?t=7744&page=70

Go to Technologies page to see whats new

LiRi

Johnny Chung Lee?  Who?  I have to admit I have no clue who this guy is?  So I Googled him and found this:

And I also found his website: Procrastineering

Mr Lee  is pretty excited about working on Project Natal and had this to say on his blog:

Xbox and Microsoft deserve an enormous amount of credit for taking on such an ambitious project. It’s one thing to say “Wouldn’t it be cool if…”, but it’s another thing entirely to say, “let’s dedicate the resources to really make it happen inventing whatever needs to be invented along the way.” I have to say it’s pretty neat building the future.

It looks like Project Natal is in some pretty capable hands.  The more I see and read about it, the more I want it.

Durante esta semana se llevó a cabo la E3 (Expo Entertainment Electronic) en Los Angeles. En ella se presentaron las empresas más importantes de videojuegos, las cuales mostraron sus productos para los próximos meses. Entre las distintas conferencias aquella que más destacó este año fue la de Microsoft con su “Project Natal” (Proyecto Natal).

Project Natal

Este dispositivo corresponde a un aparato (ver imagen de arriba) que se coloca en la parte inferior o superior de una TV, muy similar al sensor de la Nintendo Wii, la diferencia principal radica en que este no necesita de un control o mando para realizar acciones, para esto utilizamos nuestro cuerpo, como se ha podido apreciar en los distintos videos presentados en esta E3.

Primer video (Announcement Trailer) presentado en la conferencia el día Lunes 1 de Junio

Microsoft, por su parte ha explicado su funcionamiento de la siguiente manera:

Es compatible con cualquier Xbox 360, y el Project Natal es el primer aparato del mundo que combina una cámara RGB, un sensor de profundidad, un micrófono multi-array y un procesador dedicado corriendo software en un dispositivo. Al contrario de lo que hacen las cámaras 2-D y los controladores, Project Natal es capaz de capturar el movimiento de tu cuerpo de forma tridimensional, mientras responde a los comandos y directrices de tu voz. Además, el sensor de Project Natal no es dependiente de la luz. Te puede reconocer simplemente mirándote a la cara, y no reacciona sólo ante simples palabras, puede entender lo que dices. Llama a un jugador en un juego de fútbol y te responderá.

En otras palabras, este dispositivo está dotado de una cámara, sensores y un micrófono, que son usados para capturar de forma tridimensional todo lo que ocurre delante de él y reproducir de una forma fiel los movimientos del jugador en la pantalla.

Con el pasar de la semana, y la mayor parte de la web hablando sobre este dispositivo revolucionario, se reveló un detalle que muchos consideraron una bomba para Nintendo, la compañía pionera en el mundo de los videojuegos interactivos.

Johnny Chung Lee

A finales del 2007, Johnny Chung Lee se hizo famoso en Internet por sus modificaciones al Wii Mote para distintos propósitos, entre ellos el Head Tracking (seguimiento de la cabeza) o Finger Tracking (seguimiento de los dedos), este último muy similar a lo que se hace en los dispositivos táctiles sobre sus superficies (finger gestures). Todo esto, de manera independiente, en aquel entonces muchos nos cuestionamos el por qué Nintendo no contrataba a una persona como él, quien tenía mucho que aportar.

Con el paso del tiempo supimos que Lee estaba trabajando con Electronic Arts y durante esta semana nos informaron que uno de los responsables detrás del concepto del Proyecto Natal fue Johnny Chung Lee (¿qué pasó Nintendo?).

Pero eso no fue suficiente, con el pasar de los días Microsoft mostró un video en la E3 que nos dejaba asombrados con lo que puede ofrecer el proyecto Natal. Kate, una joven, interactúa con un niño virtual llamado Milo. Ambos hacían gala de lo que se pretende obtener con este proyecto. La persona encargada de esto es Peter Molyneux.

Milo & Kate interactuando a través de una conversación

Según las personas que tuvieron la oportunidad de interactuar con “Milo”, nos cuestan que es capaz de seguirnos con su mirada si nos movemos, responder a nuestros comentarios, alegrarse si le alabamos. Milo va mucho más alla, por ejemplo, de un Tamagotchi, debido a que se crea un un vínculo más cercano y significatico por medio de la conversación e interacción con este niño.

El proyecto natal sugiere una revolución y un cambio abrumador en lo que hoy conocemos como “interactividad”. Aunque personalmente creo que esté dispositivo aún se encuentra en vías de desarrollo y que su implementación debe tomar un tiempo prudente para llevar a nuestros hogares lo que se está mostrando aquí.

Como ejemplo, podemos mencionar el juego de Nintendo Wii Red Steel con el cual aprendimos que antes de encantarse con videos que prometen una interacción 1:1 es necesario probarlo con nuestras propias manos.

Cabe destacar en este punto, que yo aún estoy en contra de un dispositivo como este para los videojuegos, estoy acostumbrado a usar un joystick, control, etc. De hecho, se me hace dificil jugar en mi iPod Touch juegos como Sonic 2 o Chippy, ya que se pierde el sentido primordial a la hora de jugar…el tacto. Otro punto a considerar es el precio, ¿cuánto estaremos dispuestos a pagar por este aparto?.

El proyecto natal es un concepto que se puede aprovechar en muchísimas otras áreas, como la gestión o presentación de información. De hecho estoy seguro de que esto es el siguiente paso, o sucesor, de las actuales pantallas tactiles y en las mismas palabras de Peter Molyneux…”This works today, now” (esto funciona hoy, ahora), pero aún necesita tiempo, para que tengamos en nuestras manos una tecnologíafuncional como esta, muy similar a la que se muestra en la película “Minority Report“.

I’ve just heard Johnny Lee talking through his student day inventions – hacking together a projector, a Wii remote and his laptop to create a low cost interactive white board (for sub £50), a 3-D screen experience and more.

Check out his videos – 12m Youtube views, so almost as popular as Lauren Luke…

The big thing for the future is to create sensory environments – flat screens that feel like you are pushing buttons or creating physical sensations of touch to recreate objects virtually. Hmmm.

El mando de la Wii esconde muchas cosejas.

Hoy mencionaremos el detector de infrarrojos que le sirve a este simpático muchacho, Johnny Chung Lee, para hacer unos curiosos y muy baratos cachivaches.

El primero: Detecta la posición de tus dedos y puedes manipular tu ordenador como Tom Cruise en Minority Report… bueno, un poquín.

El segundo: Una pizarra digital, plenamente funcional. Sí, sí… como esas que valen 600 o 700 eurazos.

El tercero: Un aplique para que el ordenador detecte la posición de tu cabeza y genere las imágenes acorde con eso… Es como mirar un holograma… ay, que recuerdos

Aquí os dejo en enlace donde explica sus proyectos con el mando de la Wii

http://johnnylee.net/projects/wii/

¡Es un figura!

This is rather old, but I keep finding out that people I know have never seen it, so…

It really is good. 6.8 million viewers, much less me, can’t be wrong

At times, over the last two decades, we fall into a trap to believe that the major tech mountains have all been climbed, and that new thinking will be more of the incremental, not disruptive variety.

I am not sure why that happens – maybe, in our real-time world, where we share and publish judgements and observations an order of magnitude more frequently than in decades past, it takes more to get us to lift our heads up to observe true disruptive innovation.

That’s why I treasure when things like the iPhone and its ecosystem happen to fundamentally shake things up.

I felt the same excitement last week, when, at the TED conference, a team from MIT demonstrated a new approach for contextual environment augmentation, using about $300 of hardware and their invaluable collective smarts.

To start the dialog, take a look at one of the sample videos that the MIT team came up with.

The main components of what they demoed that I love included:

• The gesture interface, from using your hands to frame a photo, to the multi-touch implementation
• The way it addresses that “What’s in focus” issue, by using you fundamental positioning and viewplane to augment/complement what you are viewing
• How it employs text recognition within an image to further understand context
• The hardware the team employed, weighting in at a $300 pricepoint!
• The use of the physical world (the paper I am reading, the photo I am viewing, etc.) as the actual display, which was mind-blowing (e.g., seeing my real-time flight status displayed on the ticket I am holding).
• My favorite of all their examples (albeit impractical): Projecting the tag cloud of information about the person I am speaking to ONTO the person as we speak

Augmenting the physical object I am interacting with information with is one of those simple, but great ideas (done before, but here it has two attributes lacking in prior efforts:  portability and affordability)

There are other interesting interface options being worked on in parallel.  Some examples include:

• Smart clothing, incorporating passive sensors and active controls
• Heads-up displays in eyewear – Note:  Still LOTS of work still to do here to avoid looking Borg-like
• Contextual extension of existing devices (GPS, Accelerometers, Compass) – all now minimum-to-ship requirements for the smartphone players nowadays
• Flexible OLED display advances
• Pen-based interfaces (still plugging away after 8 plus years).

There are lot of blades in the Swiss Army knife of mobile interface options – the art is determining how they are creatively and cost effectively deployed in creating new user experiences.  A great example is the Augmented Realty for the Blind on the Android G1 Platform.

For me, while we have lots of efforts to re-create fixed location Minority Report solutions (from the Microsoft Surface to Johnny Chug Lee’s fun work while he was at CMU), the iPhone has shown that the next pervasive shift in man machine interface is all about mobile, and with great ideas like the one demonstrated by the MIT team last week, I can’t wait to see what will happen next!

Johnny Chung Lee is the master of the YouTube instructional video. In fact, Johnny’s “Head Tracking for VR using the Wii Remote” was nominated for best instructional video of 2007.

Johnny has tons of awesome videos about using the wii-mote for motion tracking. There are videos and source-code for head, finger and pen tracking enabling VR and multi-touch surfaces for the hobbyist. He has inspired hundreds of home-made VR projects and even has a forum. His blog also provides up-to-date information on his tinkerings.

Johnny’s lesser known work involves automated projector calibration using embedded light sensors and actually comprises his doctoral thesis at CMU. Instead of using high cost camera based tracking systems, Johnny presents a method for using low cost projectors for simultaneous tracking and content display. Essentially, light sensors on the object that is being tracked deduce their own position given a pattern that is being projected.

Black and white gray code projections are used to determine the pixel position of each sensor in reference to the projector. When this information is paired with a digital model of the object and the sensor positions, a system of equations solves for the parameters of the projector and the object. The solution yields the intrinsic parameters of the projector lens and the 3d position and orientation of the object.

Johnny demonstrated what interactive projector based tracking would look like in the video below (and a corresponding UIST 2008 paper). Note that the hybrid infrared/visible light projection in this video was faked using the wii-mote (and thus can only track 4 points).

I will be posting an in-depth explanation of this process soon. Hopefully with some nice videos.

You probably have already seen this video of Johnny Chung Lee and his Wii; if not, you should.

This is someone that has figured out how to do what he loves (and has found an audience that appreciates his work). After seeing one of his new projects today, I took a look at his CV. Turns out that this Carnegie Mellon PhD has not only created a following on YouTube (Top Rated Video, Dec 2007), but he also was the Project Creator on:

A large scale community art project for the City of Pittsburgh for the urban renewal initiative of the East Liberty area. The project involved the construction of a 45 foot tall slingshot to launch paint-filled balloons fired by community to temporarily transform a condemned apartment building into an 18-story tall piece of artwork.

There are a few lessons here—and I wonder if it is a coincidence that another life lover and Internet sensation, Randy Pausch, also claimed this university—in the same department and field of study even?

Is there something we ought to learn more about when we discuss human interaction (the area of study for both these individuals)?

I think so.

Para os que se interessaram em montar a caneta, abaixo segue o esquema do complexo circuito montado! =D

Esquema simples

Falou!

Não é por nada que o vídeo de Johnny Chung Lee tá na categoria de “Rated most jaw-dropping” lá no TED.com. O cara botou muito quente em superutilizar um “mero” (há! de mero não tem nada) controle de Nintendo Wii para fazer uma coisa totalmente diferente do propósito do joystick. Seguindo os passos do nosso mestre inspirador, construí essas canetinhas cuja ponta é um LED infravermelho.

Johnny Chung Lee figured out a way to turn a normal computer display into a close approximation of a virtual reality screen.  He also knows how to create an interactive whiteboard – which typically cost more than $1,000 – for $60 in parts and a little software code.

So what does he do with this information?  He shares it.  His video about a computer display has been viewed more than 6 million times on YouTube.  More than 700,000 people have downloaded the software for making an interactive whiteboard, and a “Lego robotics club” of 5^th graders (isn’t it great that such a thing exists?) built one in 4 hours.

Before you finish reading this post, do yourself a favor and DON’T file this away into the “computer geek” section of your brain (along with Linux, etc.).  Why can’t (and shouldn’t) this be the approach for all great innovation — especially in the nonprofit sector where resources are scarce?

Mr. Lee is quoted in yesterday’s NY Times asking, ‘”Would providing 80 percent of the capability at 1 percent of the cost be valuable to someone?” If the answer is yes, he says, pay attention.’

Imagine if every great innovation in fighting poverty, in hospital administration, in public schooling or early childhood education were open-sourced.  Wouldn’t we all be a lot smarter and more likely to focus our attention on models that work?

I posted a manifesto the other day making the same point as Mr. Lee: that the innovation, the insight, and the direct impact matter; and that sharing that innovation with others matters just as much.

I think I’m posting one of Johnny Chung Lee’s lines at my desk: “If you create something but nobody knows, it’s as if it never happened.”  Get out there and spread the word.

(And if you have 5 minutes to be amazed, watch Johnny Chung Lee’s demonstrations on the TED website).