It is getting increasingly common to find voice recognition features in gadgets. From virtual assistants to keyboard input options, even smartwatches are now equipped with the feature. However, limitations like ambient noise make it less effective as the feature does not get turned on when we call it out sometimes. Also, the application gets worked up in meetings and similar settings and gets activated.
Good news, though. There is a perfect replacement of the conventional voice-recognition software in the making.
Scientists at the Seoul-based Pohang University of Science and Technology have developed a sensor that can accurately recognize your voice. Its USP is that it can pick up skin vibrations directly and decode them to produce precise results!
Unlike regular voice recognition applications that rely on sound pressure to recognize voice easily dampened by ambient noise, the sensor is designed to receive body vibrations, which are not affected by any surrounding sounds. The sensor is super sensitive and does not require you to speak loudly in noisy settings, in an effort to activate it. Even soft whispers can be received by the responsive sensor. The unique device is made up of an ultra-thin polymer film and a diaphragm having tiny holes and senses voice quantitatively, capable of hearing sounds as low as 40 to 70 dBSPL.
The sensor is flexible enough to be attached around your neck. The result is clear, distortion-free voice being received by the sensor, with increased command effectivity in a low volume. The application works even if a mouth mask is work while speaking!
There are various voice-recognition applications such as electronic skin, wearable vocal healthcare monitors and other human-machine interface that can be developed by extending this research. Quantitative reception of sound vibrations is the future of voice recognition!
The research can be further extended to various voice-recognition applications such as an electronic skin, human-machine interface, wearable vocal healthcare monitoring device.
“This research is very meaningful in a way that it developed a new voice-recognition system which can quantitively sense and analyze voice and is not affected by the surroundings,” said Kilwon Cho, a professor at POSTECH.
“It took a step forward from the conventional voice-recognition system that could only recognise voice qualitatively,” Cho said.