The company is launching an update to its gesture recognition software that it claims is 98-percent accurate under optimal conditions. August 1, 2013 6:01 AM PDT Gesture control may get a boost in the fall holiday buying season if PointGrab can convince more device makers and app developers to adopt its Hybrid Action Recognition technology. The company released a new version of its gesture recognition software, which works with the standard 2D camera on most consumer devices. The major change that could make gesture control more mainstream is the increased accuracy of the software. The company claims that it is 98 percent accurate in "optimal" conditions. The technology can tell the difference between intentional movement, such as putting your fingers to your lips to mute a TV or computer, and unintentional gestures, such as scratching your lip, according to Assaf Gad, vice president of marketing and product for PointGrab. "We don't want to teach people how to interact with devices. The devices have to be smart enough to analyze the body language of user and act accordingly," Gad said. PointGrab has updated its machine learning algorithms to accommodate more environmental factors, such as different lighting conditions and hand sizes, to improve accuracy. In countries with heavy cigarette smoking, such as China and Korea, the software needs to account for clouds of smoke in calculating hand shape and motion accurately. PointGrab has a range of 3 to 22 feet for larger screens and 3 inches to almost 20 feet for mobile devices, the company said. An earlier version of its software has shipped on about 10 million devices, including Samsung's Smart TVs as well Acer and Fujitsu PCs. The newer version started shipping on Samsung Smart TVs in mid-May, and an updated solution for PCs with Windows 8.1 TV manufacturers will ship by the beginning of the fourth quarter, Gad said. PointGrab doesn't add much cost to a device, charging partners a few dollars per system for its software. PointGrab also been integrated into CamMe, an iOS app lets you take pictures on your smartphone from a distance using gestures to activate the camera. CamMe has been downloaded about 600,000 times from the App Store, Gad said. An software development kit for app builders is available this week. Based in Israel, PointGrab has been working on gesture recognition interfaces since 2008. Two other Israeli companies, Eyesight and Extreme Reality, are also developing gesture and body movement recognition software platforms that work with a variety of devices equipped with standard 2D cameras. In addition, Leap Motion just introduced a hardware controller for enabling gesture control, and Microsoft's Xbox One will enable control of more than games with its updated Kinect technology. "I see it as good sign that we are in the right place to change way people interact," Gad said. "In the near future, we'll see more solutions that combine voice and gesture, as well as 3D.

Posted by : Unknown Thursday, August 1, 2013

The company is launching an update to its gesture recognition software that it claims is 98-percent accurate under optimal conditions.



August 1, 2013 6:01 AM PDT




Gesture control may get a boost in the fall holiday buying season if PointGrab can convince more device makers and app developers to adopt its Hybrid Action Recognition technology. The company released a new version of its gesture recognition software, which works with the standard 2D camera on most consumer devices.


The major change that could make gesture control more mainstream is the increased accuracy of the software. The company claims that it is 98 percent accurate in "optimal" conditions. The technology can tell the difference between intentional movement, such as putting your fingers to your lips to mute a TV or computer, and unintentional gestures, such as scratching your lip, according to Assaf Gad, vice president of marketing and product for PointGrab.


"We don't want to teach people how to interact with devices. The devices have to be smart enough to analyze the body language of user and act accordingly," Gad said. PointGrab has updated its machine learning algorithms to accommodate more environmental factors, such as different lighting conditions and hand sizes, to improve accuracy. In countries with heavy cigarette smoking, such as China and Korea, the software needs to account for clouds of smoke in calculating hand shape and motion accurately.


PointGrab has a range of 3 to 22 feet for larger screens and 3 inches to almost 20 feet for mobile devices, the company said. An earlier version of its software has shipped on about 10 million devices, including Samsung's Smart TVs as well Acer and Fujitsu PCs. The newer version started shipping on Samsung Smart TVs in mid-May, and an updated solution for PCs with Windows 8.1 TV manufacturers will ship by the beginning of the fourth quarter, Gad said. PointGrab doesn't add much cost to a device, charging partners a few dollars per system for its software.


PointGrab also been integrated into CamMe, an iOS app lets you take pictures on your smartphone from a distance using gestures to activate the camera. CamMe has been downloaded about 600,000 times from the App Store, Gad said. An software development kit for app builders is available this week.


Based in Israel, PointGrab has been working on gesture recognition interfaces since 2008. Two other Israeli companies, Eyesight and Extreme Reality, are also developing gesture and body movement recognition software platforms that work with a variety of devices equipped with standard 2D cameras. In addition, Leap Motion just introduced a hardware controller for enabling gesture control, and Microsoft's Xbox One will enable control of more than games with its updated Kinect technology.


"I see it as good sign that we are in the right place to change way people interact," Gad said. "In the near future, we'll see more solutions that combine voice and gesture, as well as 3D.



Translate

Like fanpage

Popular Post

Blog Archive

Powered by Blogger.

- Copyright © News and design logo -Metrominimalist- Powered by Blogger - Designed by Johanes Djogan -