Hitachi allows for more accurate multilingual speech recognition

<p></p>

Japanese technology corporation Hitachi Group announced that it has developed a speech signal processing technology that will allow for more accurate multilingual translation in noisy environments.

The technology will be able to accurately recognize speech in environments with noise levels up to 70 decibels, or about as loud as the average vacuum cleaner. Hitachi says that it works by automatically removing background noise, thereby isolating the speaker’s voice.

Users will also be able to translate their speech without needing to press a button, according to Phys.org. The goal was to develop an application that allowed for hands-free translation. It also runs in the cloud, taking up less storage data on users’ devices.

This advance in technology points toward the growing need for translation services in tourist and business destinations such as Tokyo, where Hitachi is headquartered. The company’s main focus in developing the application is to provide technology that supports Japan’s hospitality industries.

-----

Source: Phys.org

Image: PRWeb

Comments