Archive for June, 2010

FASHION crime it may be, but a multicoloured dayglo glove could bringMinority Report-style computing to your home PC.

Interest in so-called gesture-based computingMovie Camera has been stoked by the forthcoming launch of gaming systems from Microsoft and SonyMovie Camera that will track the movements of players’ bodies and replicate them on screen. But an off-the-shelf system that can follow delicate hand movements in three dimensions to manipulate virtual objects remains tantalisingly beyond reach.

The problem with systems such as Microsoft’s Project Natal for the Xbox is that they do not focus on the detailed movement of hands, limiting the degree to which players can manipulate virtual objects, says Javier Romero, a computer-vision researcher at the Royal Institute of Technology in Stockholm, Sweden. Arm movements can be captured but more subtle pinches or twists of the wrists may be missed.

Until now, capturing detail required expensive motion-capture systems like those used for Hollywood’s special-effects fests. These utilise markers placed around the body, or sensor-studded data gloves in which flexible sensors detect joint movements. “Really accurate gloves cost up to $20,000 and are a little unwieldy to wear,” says Robert Wang, a computer scientist at the Massachusetts Institute of Technology’s Artificial Intelligence Lab.

Wang has developed a system that could bring gesture-based computing to the masses and it requires nothing more than a pair of multicoloured latex gloves, a webcam and a laptop .

Hands where I can see them (Image: CSAIL)

The key to the system is the gloves, each of which is comprised of 20 patches of 10 different colours – the maximum number a typical webcam can effectively distinguish between. The patches are arranged to maintain the best possible separation of colours. For example, the fingertips and the palm, which would frequently collide in natural hand gestures, are coloured differently.

The upshot is that when a webcam is used to track a glove-clad hand, the system can identify each finger’s location and distinguish between the front and the back of the hand. “It makes the computer’s life easier,” says Wang.

Once the system has calculated the position of the hand, it searches a database containing 100,000 images of gloved hands in a variety of positions. “If you have more images than that it slows the computer down, and if you have fewer then you don’t provide an adequate representation of all the positions the hand can be in,” Wang explains.

Once it finds a match it displays it on screen. The process is repeated several times per second, enabling the system to recreate gestures in real time.

Wang presented some early-stage research at last year’s SIGGRAPH meeting in New Orleans, Louisiana. “Back then it only worked in windowless rooms and took half an hour to calibrate,” says Wang. Now it can be calibrated in 3 seconds, he says.

Wang has already shown that the system can correctly replicate most of the letters of the American Sign Language alphabet, although those that require rapid motion (J and Z) or involve the thumb (E, M, N, S and T) have yet to be perfected.

The gloves are so cheap to make – costing about a dollar – that they could bring gesture-based computing to a wider audience, says Douglas Lanman, an expert in human-computer interaction at Brown University, Providence, Rhode Island. But if it’s going to have truly widespread appeal, it will need to lose the gloves. “Wearing a glove is an inconvenience,” he says. “Markerless motion-capture is where I think the field is moving, and where the larger commercial market will be.”

Last month, at the IEEE International Conference on Robotics and Automation in Anchorage, Alaska, Romero and his colleague Danica Kragicdemonstrated how markerless motion-capture may be possible. Their system also uses a webcam and a database of hand positions to recreate an on-screen version, but attempts to pick out a bare hand in a stream of video from a webcam by detecting flesh colours. If you reach down and pick up a ball, say, the program will aim to find a matching image in its database of the positions the had adopts as it reaches down and picks up a spherical object.

Identifying a hand using skin colour is far more difficult than picking out a multicoloured glove. Even once a hand is detected, it is a massive challenge to accurately identify its position – especially if it is holding something, says Kragic. “The object blocks out parts of the hand, preventing the computer from knowing what the hidden bit is doing.”

To tackle the problem, Romero and Kragic created a reference database containing images of hands picking up 33 different objects, such as a ball or a cylinder. They then set up a webcam, which captured 10 frames per second, and tested their system’s capabilities by filming people grasping a cup, a ball or a pair of pliers. The database had images of a hand picking up a ball, but nothing for a cup or pliers. The system successfully created virtual representation of a hand grabbing a ball, and came as close as it could to the cup by displaying a hand grasping a cylinder. It came up empty with the pliers.

These are exciting results nonetheless, says Romero, because they show that the system can not only reconstruct the gestures of empty hands, but can also generalise when dealing with some unknown objects. The shape of the pliers, and the grasp used to pick them up were too different from anything in the database for the system to find a match, but by expanding the reference database it should be possible to overcome that, he says.

To make identification faster, Romero has incorporated an algorithm to rule out unlikely hand positions based on previous estimates of hand pose. For example, if the last hand position was a hand stretched out with splayed fingers, the algorithm rules out database images of hands that are clenched into a fist. While this helps the system operate in real time, it creates problems of its own: if the hand moves very fast, it can indeed “jump” from being splayed out to being clenched. In this situation, Romero’s set-up struggles because that algorithm will rule out the correct pose.

Romero claims the system is already attracting interest from makers of prosthetics, who want to improve their understanding about how people grip objects.

It will also, of course, interest game makers, says Takaaki Shiratori of the Robotics Institute at Carnegie Mellon University in Pittsburgh, Pennsylvania.

COURTESY: http://www.newscientist.com


Read Full Post »

Christian Kandlbauer mind-controlled arm prosthesis

A mind-controlled robotic arm is bringing new independence to an Austrian man who lost his arms, even allowing him to drive a car, the arm’s maker announced earlier this week. Christian Kandlbauer passed his driving test “with flying colors” using a mind-controlled arm prosthesis. Created by the German firm Otto Bock HealthCare, the arm has allowed 23-year-old Christian Kandlbauer to pass a driving test with flying colors, giving him the freedom to drive the seven kilometers to and from work each day without assistance. The device is the first mind-controlled arm prosthesis in Europe. After a high-voltage electrical accident struck Kandlbauer back in 2005, both his arms had to be amputated. His right arm has since been replaced by a myoelectric DynamicArm, which now operates the steering wheel when he drives his specially equipped Subaru Impreza, modified for him by Paravan. Taking the place of his left arm, however, is the new, seven-jointed prosthetic arm, which he controls in real time using just his thoughts. He now operates that arm to control the direction indicator lights, horn and all other safety-related devices of his vehicle.

Targeted Muscle Reinnervation

The prosthesis is based on targeted muscle reinnervation (TMR), a process by which four key arm nerves are systematically relocated to the surgically segmented chest musculature. Once reinnervation is complete — it can take several months — the surface of the chest forms an interface to the patient’s brain. Electric nerve impulses coming from this region can actually be sampled and processed by surface electrodes, according to Otto Bock. Now, powerful micro-controllers in the prosthesis calculate the motor commands underlying the impulses in real time and generate the equivalent control commands for the arm. For Kandlbauer, who works in a garage, it’s been a dream come true. He obtained his driver’s license in October and has been traveling back and forth to work in his own vehicle ever since.

‘An Important Extension’

“What’s remarkable about the Otto Bock prosthetic technology is the way they have re-engineered the nervous system to make the prosthesis more useful,” James Cavuoto, editor of Neurotech Reports, told TechNewsWorld. “Innovative strategies like this will likely grow the market for motor prostheses in the years ahead,” Cavuoto added. Indeed, following 10 years of pioneering work in TMR, “it is exciting to see this promising new source of command signals being applied to commercial, clinical prostheses,” agreed Gerald Loeb, professor of biomedical engineering at the University of Southern California. It’s not clear whether it’s truly a case of “mind-control,” however, “particularly because other researchers have been working — with much less success — on reading out command signals directly from the brain, where the term seems more appropriate,” Loeb told  Nevertheless, TMR is “an important extension of the original concept of myoelectric control because it enables users to operate multiple degrees of freedom in a way that is much more natural and intuitive than has been possible before,” he explained. “That sort of command capability is essential to take advantage of the much more sophisticated mechatronic hands and arms that are being developed by several research and commercial groups, including Otto Bock,” Loeb added.

A High-Tech Index Finger

Otto Bock is also working on a “high-tech index finger” for Kandlbauer that can sense temperature, identify rough and smooth surfaces and feel the strength of a handshake. “Micro-sensors that record the temperature, gripping strength and surface characteristics of the object being gripped are integrated into the tip of the index finger on the prosthetic hand,” explained Hubert Egger, head of the mind-controlled arm project. Currently, however, the prosthesis with sensory perception is only a prototype. It will likely take another four years of development before Kandlbauer can use it on a day-to-day basis, Otto Bock said.


Read Full Post »

4G also called as Fourth-Generation Communications System, is a term used to describe the next step in wireless communications. A 4G system can provide a comprehensive IP solution where voice, data and streamed multimedia can be provided to users on an “Anytime, Anywhere” basis. The data transfer rates are also much higher than previous generations.

3g-to-4g14G is the latest phenomenon to metamorphose the cellular superhighway, soon to succeed its predecessors 3G and 2G.G stands for the generation in mobile technology that have come of age maturing down the line over 20 to 30 years. The first generation mobile technology were based on analog telecommunication standards. 2G technology evolved in 1990s, introducing digital circuit switched transmission and SMS text messaging. This was followed by a big leap in the form of 3G, aka IMT-2000 that marked the major advances in mobile technology including efficient data handling enabling video calls, accessing websites in oiginal format and much more. For 3G what better example can you have if not iPhone 3G. 4G system is anticipated to revolutionize the functioning of communication networks with comprehensive and secure IP based solution facilitating voice, data and streamed multimedia Anytime, Anywhere and with much higher data transfer rates than its predecessors.

4G- Blueprint of Goals

4G mobile is under the process of being established according to the set standards. It’s features are more of a blueprint based on goals rather than requirements.  Let’s take a snap of the goals set for 4G

  • Faster data transfer rates
  • Enhanced security measures than its predecessors
  • Reducing blips in transmission when a device is moved from one networks coverage to another
  • 4G mobile networks should also use a network based on the IP address system that is used for Internet

One World One device: Global roaming in 4G


The fourth generation would completely transform the mobile phone networks, in to end to end IP based networks, couple this with the arrival of IPv6.  With a unique IP address for all devices in the world, it would initialize full IP based communications from a mobile device right from the core of web.

It would truly harmonize global roaming, super high speed connectivity and transparent end user performance on every mobile communications deice in the world.

Once on the stage is set, 4G would offer 100 Mbps to roaming mobile device globally and around 1 Gbps to a stationary device. Clearly, this could have wider implications including overwhelming wireless performances like enhanced video conferencing, streaming panoramic video and much more.

The giant leap would require a massive infrastructural makeover in the phone networks – multiplying the traffic load on the internet as a whole. 4G needs swifter backbones and oceanic links for a smooth run. For paralleling things, think of 1 billion 100mb nodes attached to a network over night.

Who’s using 4G?

There are two major systems in U.S, which are using the 4G mobile technology  – WiMax, backed by Clearwire and Long Term Evolution or LTE. WiMax’s majority owner is Sprint Nextel. They had started testing services in Baltimore in 2008 and waiting to expand their coverage over 80 cities by the end of 2010.

Long Term Evolution is backed by Verizon. According to cnet, Verizon has completed initial 4G wireless test, but not available for widespread use until 2012.

WiLAN and 4G Network technologies have signed a U.S $ 31 million purchase order as a part of 5 year supply agreement to build a fourth generation cellular network throughout US

Well, 3G is yet to reach its zenith and 4G couldn’t be an overnight phenomenon. In such a situation we can just keep our fingers crossed. There’s an air of speculation that suggest the fourth G would take on the mobile world by 2010. For the time you can just wait and watch our updates.

The main objectives of 4G are:
1)4G will be a fully IP-based integrated system.

2)This will be capable of providing 100 Mbit/s and 1 Gbit/s speeds both indoors and outdoors.

3)It can provide premium quality and high security.

4)4G offer all types of services at an affordable cost.

4G is developed to provide high quality of service (QoS) and rate requirements set by forthcoming applications such as wireless broadband access, Multimedia Messaging, Video Chat, Mobile TV, High definition TV content, DVB, minimal service like voice and data, and other streaming services.

4G technology allow high-quality smooth video transmission. It will enable fast downloading of full-length songs or music pieces in real time.

The business and popularity of 4Gmobiles is predicted to be very vast. On an average, by 2009, this 4Gmobile market will be over $400B and it will dominate the wireless communications, and its converged system will replace most conventional wireless infrastructure.

Data Rates For 4G:

The downloading speed for mobile Internet connections is from 9.6 kbit/s for 2G cellular at present. However, in actual use the data rates are usually slower, especially in crowded areas, or when there is congestion in network.

4G mobile data transmission rates are planned to be up to 20 megabits per second which means that it will be about 10-20 times faster than standard ASDL services.

In terms of connection seeds, 4G will be about 200 times faster than present 2G mobile data rates, and about 10 times faster than 3G broadband mobile. 3G data rates are currently 2Mbit/sec, which is very fast compared to 2G’s 9.6Kbit/sec.

Source:wise Geek4G technology

Read Full Post »