| Anyone who saw Spielberg's «Minority Report» would hardly have guessed that the creators of this science fiction film were filming an unknown to the general public and still raw (at that time) development that had been underway for years by American computer scientists. True, the filming was not without special effects, but the idea was based on a real system.
The creators of the system that later evolved into G-Speak are John Underkoffler and his colleagues from the MIT Media Lab. It was there that Spielberg and his subordinates spied the current control of data using gestures.
Long before the «Minority Report», Underkoffler thought about how to bring computer graphics into the real world. He dreamed that pictures, diagrams, tables could be displayed on walls, floors, palms, tables and chairs.
Moreover, it would be good to control these virtual objects with simple gestures. Draw a circle with your finger — cut out an object from the frame, move your palm down — move the fragment to a new photo, make a few more passes — send the received one by e-mail. Something like that.
The conclusion was clear — projectors are needed. Moreover, scientists immediately thought about placing one or, better, several such installations on the ceiling and walls, where light bulbs are usually placed. A light bulb has a «privileged position» in the room — it can see everything well.
The system's creators joked that a light bulb is also a projector of sorts, only showing one pixel. So all that was needed was to «increase the resolution.» That's what the American researchers did, building a very small digital projector (not without the help of a number of partners from industry).
Gesture perception turned out to be a more complex task. It was decided to build a tiny camera into the same «bulb-projector». The camera would perceive hand movements (in special gloves equipped with clearly visible markers) and transmit them to the computer. Then it was a matter of technology: the PC had to recognize gestures on the fly, determine the coordinates of the hands in space and interpret them as commands. However, creating such software took a lot of time. It all looks simple only in words.
Several of these «lightbulbs» covering different «angles of fire» in a room would create the environment Underkoffler dreamed of. The authors even came up with a funny term «input-output lightbulb», although this device has nothing to do with a light bulb.
This is how the Luminous Room appeared in the 1990s at the Massachusetts Institute of Technology. Underkoffler spent many years there refining the principle of «airborne graphics control», and it was there that members of Spielberg's team spent entire days in order to later show the computer interface of the «distant future» in the film as realistically as possible. John Underkoffler, by the way, became a scientific consultant for «Minority Report». We owe him, among other things, for the thoughtfulness of the «realities of 2054».
The film, as we know, was a success (by the way, another idea from Minority Report, albeit on a different principle, was tried to be implemented in Britain — this is the prediction of the names of future murderers). But the main thing is that the “future” in this context turned out to be not so distant.
Luminous Room «pumped up its muscles», acquired a more advanced «intelligence» and turned into the G-Speak environment. For its further development and promotion to the market, a company Oblong Industries was recently created (Underkoffler became its leading scientist and one of the founders). Now the amazing interface is presented to the public under the slogan «Minority Report in reality». Advertising, where would we be without it. And the system is truly impressive.
But if you're too lazy to wave your hands for hours on end, you can try turning to Nicole. Literally: «Nicole! Launch IE and find Beethoven's «Moonlight Sonata» on the Internet for me, please.» It will launch it and find it. And then it will play it, obeying the same voice commands.
The creator of this system, a kind of virtual secretary, or more correctly, desktop personal assistant, is Tom Moir from the School of Engineering and Advanced Technology, Massey University, New Zealand. Yesterday, December 2, Tom showed the virtual assistant Nicole in Auckland, at the 15th International Conference on Mechatronics and Machine Vision M2VIP.
Brazilians consider their Nicole to be the most realistic virtual assistant in the world. The most or not, but this character can transform from a secretary into a TV presenter, and can also undress. But such an «option» is of no use in the Moira project. As well as the fact that in reality Nicole's bot is full, from head to toe, and even capable of putting on different dresses and suits, because in the guise of an assistant, the girl is displayed on the screen as a «talking head».
Of course, voice control is nothing new in the computer world. Moir's achievement is that he brought this principle, if not to perfection, then at least to a decent level of complexity, and also combined voice control with a virtual character. It should be said here that the image of Nicole, or rather, an entire bot with dozens of facial expressions, was provided to Tom by the Brazilian company Guile 3D Studio, which specializes in «virtual people».
Nicole's facial expression can reflect surprise and anger, joy and thoughtfulness, the movement of her lips is consistent with the phrases she speaks, she can cry and express doubt in the words of her interlocutor. Perhaps, her abilities are even excessive for the role of an on-screen «secretary». Yes, one of these faces is Tom Moir. Moir taught the Brazilian beauty all sorts of tricks: working with files and folders, launching selected audio and video, searching for data, answering questions like «what time is it?», reading news out loud, and the like.
But Tom did not stop at controlling only virtual objects. After all, he believes that Nicole should become an assistant not so much for lazy users as for disabled people, for whom it is difficult or even simply impossible to operate a computer. And if so, the virtual girl should «reach» real things. For example, turn on and off the light, fan or something like that by voice command. The system created by Moir allows all this.
Well, if you are tired of talking to Nicole, try the «thing interface». The «Mirror» (Mir:ror) device is produced and sold by the French company Violet. For 45 euros, the user gets the opportunity to turn everyday things into interactive objects. The system, which debuted last month, is built on the already familiar radio frequency identification (RFID) chips.
Usually, these inconspicuous radio markers are used in factories, warehouses and stores to account for and sort a large number of objects. But sometimes you can find original systems — RFID key fobs, microchips and cards in the role of train tickets, hotel room keys or systems against forgetful surgeons, for example. And still, RFID is not very closely used in everyday life. But what prospects are opening up!
So, Mir:ror is a desktop reader of special RFID stamps called Ztamps (they are included in the kit and can be purchased as needed), plus the corresponding software. These microchips can be placed anywhere. After an object equipped with such a chip is brought to the «Mirror», the computer will execute a set of commands previously «associated» with this object. This principle is very similar to the work of the «subject» system of teaching English, only the capabilities of the «Mirror» are much wider. What you program to launch this or that «software» is a matter of taste. The French offer different options: from a mug to sneakers.
The RFID chip carrier, which also allows you to run a set of commands from the «Mirror», is the «micro-rabbit» Nano:ztag. In appearance, this «beast» repeats the much larger interactive Internet rabbit Nabaztag, which earned a good reputation and fame for Violet several years earlier. It independently goes online (it perceives commands by ear) and has a voice (it reads news, exchange rates or stock market reports), movable ears and multi-colored illumination.
For example, a book placed on the «round» of the device will automatically launch the Internet and open a page of the publisher's website or a video containing additional information, or the author's website. You can get the weather forecast for tomorrow by holding an umbrella to Mir:ror; pick up a souvenir from your last trip — the computer will show the corresponding photo album; your child's favorite toy will make the PC read a fairy tale; a pack of cigarettes can be used to count the number of cigarettes smoked per day or week; an MP3 player placed on the «Mirror» will open a folder with a «strategic» stock of songs on the computer, and a mobile phone — the website of the warranty workshop. Why not? It's up to you to decide.
«Morning. You put your car keys on this mysterious device. A voice says: «It's raining today, 14 °C. You'll get to work in 15 minutes». The computer screen turns on images from webcams located along the previously planned route, while the voice reads the horoscope for the day» — this is the scenario offered by the manufacturer.
The computer can carry out a lot of instructions programmed by the user, just by «seeing in the mirror» an object associated with this set of actions. Much easier than sitting in front of the screen and sequentially opening all these applications manually. Such is «a little magic».
It seems to be essentially nothing complicated. We have already seen similar (partly) projects before. In the form of laboratory experiments and «dreams». But no one, it seems, has yet thought of releasing such a finished product for the masses. Are you still struggling with the keyboard? G-Speak, Nicole, Mir:ror offer an alternative. How convenient it is is another question. But the main thing is when there is a choice, isn't it?
Based on materials from the site http://membrana,ru |
