A new sensor: Microsoft can also play mobile phone from a distance


Microsoft researcher recently announced that they are developing a new electric field sensor, in the absence of the touch control allows the user to mobile phones. This technique has not been named, but Microsoft has to give it a definition: low cost, transparent, 3 d, interactive, and so on. Through this kind of advanced technology, when a user’s hand suspended on the phone’s screen, and can be induced. And this will be more than the samsung Galaxy series mobile phone used by the “window” function is more advanced, and more humanized.

Microsoft research, said: “we adopt thin, transparent, and designed a low cost material electric field sensor, can track 3 d finger and hand animation, can also capture a mobile device suspended gestures. Our simple electronic drive is based on a ready-made chip, does not need to build custom analog electronics. We describe our transparent electrode array design, using a machine learning algorithm to map 3 d location of receiver. Shows us the non-contact 3 d motion accurately positioning the hands and fingers.”

last year, foreign media The Verge reported that nokia’s first Windows Phone 8.1 will use some kind of gesture detection technology, but in The end not to cash, so we expect Microsoft and nokia is likely to apply this technology in its next generation of WP devices.

Microsoft seems increasingly want to get rid of the image of square, old users continuously launching new technology, especially in the operation level of PC, mobile phone. Such as the second generation of Microsoft Surface Touch Cover has supported gestures, but demand must be performed on a keyboard gestures. However, Microsoft research, according to a new project is the break through the limit, use gesture support outside of the keyboard, the new user experience is based on the physical keyboard keys between infrared sensor for identification.

Microsoft research, said: “we are developing a new type of enhanced mechanical keyboard, using the spread between the physical keyboard key matrix, the low resolution of short infrared sensor to the direct contact with the keyboard or keyboard is variety of area above the gesture recognition, the results obtained high frame rate motion data, but the overall is still rough. We improved a machine learning, on the basis of its support only static classification, add dynamic and time dimension of the gesture. We plan to adopt action signature, a classifier can be predicted using the motion history image and random technology to identify a large number of motion. At present, we have each frame capture has reached 75.6% of the accuracy of the classification. This project is not yet mature, of course, we also will be based on user feedback to improve products.”

sources: tencent digital, compile: Zhang Xiaowei