Making Computers More Human

Perceptual computing to bring natural human responses to machine interactions.

Intel Perceptual Computing Group piloting facial recognition app

The Intel Perceptual Computing Group led by Achin Bhowmik is currently piloting a facial recognition application for securely logging into computers running Windows 8.

In an attempt to unshackle people from the archaic keyboard and mouse, technology companies are bringing human-like senses to computing devices so they can understand speech, gestures and touch.

Momentum toward new computing interactions has been gathering for a decade. People began using motion sensors to interact with gaming consuls such as Microsoft’s Xbox and the Nintendo Wii. Touchscreen interfaces to control smartphones, tablets and Ultrabooks became commonplace. Voice recognition technologies such as Siri allowed people to tell their devices what to do. Yet despite these advances, the keyboard and mouse have remained ubiquitous.

“Control, Alt, Delete . . . that’s not natural,” said Mooly Eden, senior vice president and president of Intel Israel. “We should be able to communicate with a computer the same way we communicate with one another.”

Eden has charted a team within Intel, the Perceptual Computing Group, to change the human-machine interface so that working with computers becomes more friendly, humanlike and easier. The director of the year-old “startup” confronting that challenge is Achin Bhowmik, who believes that the future of personal computers depends on their vastly improved perceptual abilities, which requires moving innovation beyond the highly developed brain to more nascent sensory subsystems of computers.

“The laptop is still primitive with only one eye, one ear and they are now just getting touch,” he said. “By giving computing devices 3-D vision systems like human beings, we can bring natural interaction to PCs and open up a whole new dimension not just for PCs, but for smartphones, tablets, media boxes, vending machines, cars and almost anything that connects to the Internet.”

Perceptual Computing’s Hardware Problem

Tim Bajarin, a principal analyst at Creative Strategies who follows perceptual and wearable computing technology trends, calls the lack of progress a hardware problem. Camera and sensor technologies need to improve and drop in price before a wave of new, useful applications actually solve real problems rather than make more immersive games.

“Sensors in cameras today are not that good at capturing a lot of details,” said Bajarin. He noted that it will be another year or two before affordable sensors are good enough to handle distance and close proximity motion commands and distinguish between four or five people simultaneously interacting on the same device.

Samsung’s eye-tracking technology inside the recently released S4 smartphones is an example of what consumers can get today, according to Bajarin. Then there’s the new USB dongle by Leap Motion, which brings a 3-D gesture tracking workspace to PCs, something he says is more in line with Intel’s approach to bringing perceptual computing to laptops, desktops and bigger screened all-in-one computers.

“Leap Motion is ahead commercially with 200 apps coming based on its SDK [software developer kit], including from big-name entertainment companies such as Disney and EA,” Bajarin said.

Computers See, Hear and Speak

Bhowmik has pulled together researchers, human factors engineers, optics engineers, system integrators, software developers and designers from inside Intel and across the industry. The fast- growing group has been creating sensor technologies and advanced software algorithms, including an SDK that allows application makers to design new experiences.

Gesture control of PC in Intel Perceptual Computing lab

Yuriy Kozachuk, a marketing engineer in the Intel Perceptual Computing Group, uses a 3-D camera mounted atop an all-in-one computer screen to demonstrate hand gesture control.

Examples of the hardware innovation underway now at Intel and other companies include a 3-D visual system that recognizes the environment and gestures, a dual microphone system that tunes to its owner’s voice commands and a haptic system that senses touch and biometrics. The challenge is making all of these systems work together.

“It’s not just about 3-D cameras,” Bhowmik said. “It’s about combining speech recognition, touchscreen and algorithms to read your facial expression and give human-like senses to the computer.”

While Siri  and other voice services are processed in the cloud, Bhowmik wants to take advantage of a device’s native processing power to deliver responsive voice commands. Dragon Assistant, available today on Dell’s XPS PCs, allows users to control a PC using natural-language voice commands to perform such tasks as updating Facebook, searching and playing songs in a music collection and finding and downloading information from the Internet.

Letting People Compute on Their Own Terms

Instead of a race to create the best multi-modal experience that trumps the tired keyboard and mouse, Bajarin sees a future where people can choose how they want to communicate with each particular compute device they use.

“There will be many ways to interact with your devices, but in the future different screens will adapt to input that’s most appropriate to the user,” he said. “Today it’s still about touch and voice, and the rest, like gesture recognition, remains novel.”

Enabling devices to understand and communicate naturally with people will take lots of work, Bajarin said. “When voice helps us get things done faster by combining different controls and commands, it could be the one thing that gets the rest of perceptual computing modes into the mainstream.”

Making Computers More Human

Perceptual computing to bring natural human responses to machine interactions.

Intel Perceptual Computing Group piloting facial recognition app

The Intel Perceptual Computing Group led by Achin Bhowmik is currently piloting a facial recognition application for securely logging into computers running Windows 8.

In an attempt to unshackle people from the archaic keyboard and mouse, technology companies are bringing human-like senses to computing devices so they can understand speech, gestures and touch.

Momentum toward new computing interactions has been gathering for a decade. People began using motion sensors to interact with gaming consuls such as Microsoft’s Xbox and the Nintendo Wii. Touchscreen interfaces to control smartphones, tablets and Ultrabooks became commonplace. Voice recognition technologies such as Siri allowed people to tell their devices what to do. Yet despite these advances, the keyboard and mouse have remained ubiquitous.

“Control, Alt, Delete . . . that’s not natural,” said Mooly Eden, senior vice president and president of Intel Israel. “We should be able to communicate with a computer the same way we communicate with one another.”

Eden has charted a team within Intel, the Perceptual Computing Group, to change the human-machine interface so that working with computers becomes more friendly, humanlike and easier. The director of the year-old “startup” confronting that challenge is Achin Bhowmik, who believes that the future of personal computers depends on their vastly improved perceptual abilities, which requires moving innovation beyond the highly developed brain to more nascent sensory subsystems of computers.

“The laptop is still primitive with only one eye, one ear and they are now just getting touch,” he said. “By giving computing devices 3-D vision systems like human beings, we can bring natural interaction to PCs and open up a whole new dimension not just for PCs, but for smartphones, tablets, media boxes, vending machines, cars and almost anything that connects to the Internet.”

Perceptual Computing’s Hardware Problem

Tim Bajarin, a principal analyst at Creative Strategies who follows perceptual and wearable computing technology trends, calls the lack of progress a hardware problem. Camera and sensor technologies need to improve and drop in price before a wave of new, useful applications actually solve real problems rather than make more immersive games.

“Sensors in cameras today are not that good at capturing a lot of details,” said Bajarin. He noted that it will be another year or two before affordable sensors are good enough to handle distance and close proximity motion commands and distinguish between four or five people simultaneously interacting on the same device.

Samsung’s eye-tracking technology inside the recently released S4 smartphones is an example of what consumers can get today, according to Bajarin. Then there’s the new USB dongle by Leap Motion, which brings a 3-D gesture tracking workspace to PCs, something he says is more in line with Intel’s approach to bringing perceptual computing to laptops, desktops and bigger screened all-in-one computers.

“Leap Motion is ahead commercially with 200 apps coming based on its SDK [software developer kit], including from big-name entertainment companies such as Disney and EA,” Bajarin said.

Computers See, Hear and Speak

Bhowmik has pulled together researchers, human factors engineers, optics engineers, system integrators, software developers and designers from inside Intel and across the industry. The fast- growing group has been creating sensor technologies and advanced software algorithms, including an SDK that allows application makers to design new experiences.

Gesture control of PC in Intel Perceptual Computing lab

Yuriy Kozachuk, a marketing engineer in the Intel Perceptual Computing Group, uses a 3-D camera mounted atop an all-in-one computer screen to demonstrate hand gesture control.

Examples of the hardware innovation underway now at Intel and other companies include a 3-D visual system that recognizes the environment and gestures, a dual microphone system that tunes to its owner’s voice commands and a haptic system that senses touch and biometrics. The challenge is making all of these systems work together.

“It’s not just about 3-D cameras,” Bhowmik said. “It’s about combining speech recognition, touchscreen and algorithms to read your facial expression and give human-like senses to the computer.”

While Siri  and other voice services are processed in the cloud, Bhowmik wants to take advantage of a device’s native processing power to deliver responsive voice commands. Dragon Assistant, available today on Dell’s XPS PCs, allows users to control a PC using natural-language voice commands to perform such tasks as updating Facebook, searching and playing songs in a music collection and finding and downloading information from the Internet.

Letting People Compute on Their Own Terms

Instead of a race to create the best multi-modal experience that trumps the tired keyboard and mouse, Bajarin sees a future where people can choose how they want to communicate with each particular compute device they use.

“There will be many ways to interact with your devices, but in the future different screens will adapt to input that’s most appropriate to the user,” he said. “Today it’s still about touch and voice, and the rest, like gesture recognition, remains novel.”

Enabling devices to understand and communicate naturally with people will take lots of work, Bajarin said. “When voice helps us get things done faster by combining different controls and commands, it could be the one thing that gets the rest of perceptual computing modes into the mainstream.”