Voice & Gesture Control: Beyond Touch With Windows 8

If Intel has its way, touch will be just one of the ways to control Windows 8 ultrabooks. The chipmaker is looking to add voice commands and Kinect-like gestures for what it calls “perceptual computing.”

On Tuesday, Intel announced a partnership with Nuance Communications to put a locally hosted voice recognition app on certain Windows 8 PCs, including devices from Dell. Intel has also partnered with Creative Technology - formerly known as Creative Labs, one of the early designers of PC sound cards - to design a Kinect-like camera that could eventually be integrated into the notebooks just as Webcams are now a standard feature for portable computers.

To kick things off, Intel also announced a software development kit to enable software developers to start creating applications that can take advantage of the new interfaces. Features include 2D/3D object tracking to enable augmented reality applications and even a sort of facial recognition to identify and personalize applications to specific users.

Perceptual Computing

“Your gestures, face, voice - all of those provide a more interactive, immersive experience, said Kirk Skaugen, vice president and general manager of the PC Client Group at Intel, speaking at the Intel Developer Forum in San Francisco.

(For more on the Intel Developer Forum, see Intel Dabbles In Science Fiction and Is This The World’s Smartest Coke Machine?)

Windows 8 PCs will be convertible, detachable and swivel-able. “The industry is going to bring more hardware innovation in the next 12 months than I’ve seen in 20 years at Intel,” Skaugen said.

The move beyond touch is a revival of sorts for voice control and dictation on the PC, which sprang into the market years ago with Nuance, Dragon Systems, and even Microsoft’s built-in voice commands. But the technology essentially fizzled, as users chose to use simpler, faster keyboard shortcuts and simply learned to type faster. Although Skaugen didn’t call them out specifically, the resurgence of voice control in Apple’s Siri and Google Android’s voice commands are clearly helping spark interest in voice control.

Meanwhile, the Interactive Gesture Camera from Creative, a development platform, will track the user’s eyes, nose, and mouth, and be able to detect smiles and other moods. Skaugen said the 720p camera will cost about $149. Eventually, according to Intel’s PC processor chief Dadi Perlmutter, the technology will be integrated into a standard notebook.

The camera, which actually looks a bit ugly squatting on top of a monitor, will track the user’s face, identifying the user and possibly limiting his access. That’s important, Skaugen said, as he described handing over his tablet to his young son. “I hand him the tablet to play Angry Birds to get ten minutes of shuteye, and the next thing I know he’s emailing my CEO.”

The camera will also be able to identify a user’s hands and fingers, sensing the presence of the user from 6 inches to a depth of three feet. That’s ideal, Skaugen said, for kitchen applications where a user doesn’t want to get his touchscreen monitor greasy. The camera will be able to detect gestures and hand positions, such as a thumbs-up gesture. Intel also claimed that its software development kit or SDK will be able to superimpose 3D virtual objects onto a person, such as a pair of glasses or a virtual hat.

"80% of communication is non-verbal, so smiles, blinks, gestures and all these sort of things make communication more immersive," Skaugen said.

Skaugen showed off two example prototype apps that took advantage of the camera: one, a virtual solar system, "exploded" into view when the user spread his hands. The other, a cute kung fu game, challenged a user to block patty-cake-styled "attacks" from an animated squirrel.

What’s Behind The New Interfaces?

"We've had voice and gesture for some time, but it hasn't really worked,” said Martin Reynolds, a vice president and fellow with Gartner. “Part of the reason is the software and part of the reason is the peripherals, and enough processing power to make it run. What we're seeing here is Intel attacking on this on all three fronts: developing the software, bumping the performance and working on the peripherals. Now how long that will take to create great results? Not clear. But this is clearly a shift." 

“This is a long-term initiative,” added Patrick Moorhead, principal analyst at Moor Insights. “It brings in two industry buzzwords: the Internet of things, and also natural user interface. But the essence of it is new ways to interact with your PC. So voice and machine vision. What’s going to happen is that the computer will understand context, it will understand everything that’s going on around you and interact with you and with other devices in new ways. That’s really the big picture.”

Perhaps influenced by the consumerization of IT, where iPhones and Gmail are invading the data center, or simply by the realization that you can surf the Web just fine on a three-year-old PC, Intel has ditched the “speeds and feeds” discussions of previous years - a decision that an Intel source said prompted some grumbling among the company's more traditional, engineering-oriented employees.  Granted, the company did announce its fourth-generation of Core microprocessors, code-named “Haswell.” But Intel completely glossed over mentions of core voltages, cache sizes and even clock speeds. Today’s watchword is “capabilities.”

Intel is even working with Sony on the Tap 20 - an all-in-one PC that’s resembles a giant tablet. (Think of it as a desktop PC with batteries.)

Overall, though, Intel continues to focus on ultrabooks - MacBook Air-like thin-and-light notebooks that Intel claims could comprise 40% of all consumer PCs sold this holiday season. By the end of the year, more than 70 different ultrabook designs will be on the market, Intel predicts, many of them equipped with touch capability, and priced starting at about $699. Apparently, Intel sees ultrabooks as a perfect delivery platform for perceptual computing.

 

Lead image courtesy of Shutterstock. Other images by Mark Hachman.