Gloves smart translate sign language in real time

The idea of translation, which can be written or spoken, is to facilitate dialogue for those who communicate using sign.

Social inclusion has been fueled by technology. This year, for example, Facebook has provided a function that describes the photos for the blind, and rolled a workshop to help children create prosthetics using 3D printers.

Now it's time for people with hearing difficulties or speech, using sign language for communication. With SignAloud gloves, the signs are recognized and automatically translated into spoken language and escrita- in real time, like any other online translator.

The SignAloud project - which can be translated as "signals loud" - has a simple operation: gloves have supersensitive motion sensors, which capture the different gestures made by the person. The data obtained here is sent via Bluetooth to a computer which analyzes the signals from a database.

When the machine is a corresponding signal, it translates into words that appear on the screen and are "spoken" by the computer in a similar speed to that of a normal conversation. In addition to useful, the gloves are practical, discreet and comfortable, and can be used in day to day without looking like a weird robotic suit.

The SignAloud were created by students Navid azodi and Thomas Pryor, the University of Washington, who used the free time between classes and equipment provided by the college to take the project forward. The effort paid off: they were first in the Lemelson Prize, the MIT, which awards the best technological inventions of US youth. Azodi and Pryor received $ 10,000 to improve the project - and 15 000 each.

The money, say the two, will be used to adapt the gloves to other languages ​​in the world signals, since the prototype that won the Lemelson students only works for American Sign Language, ASL, and each country has its own language correspondent - the Brazil, for example, is called Libras (Brazilian Sign Language).

Young people also plan to improve the database, reducing context errors - you know, like those that happen when Google translates literally one sentence and the meaning is weird. Another idea is that SignAloud can be used in hospitals to monitor movements of patients who have suffered a stroke or an accident that harms the movement.