Fujitsu develops hands-free speech translator
Hospital workers often have their hands full with tasks such as patient care, plus they also often have to communicate with patients who speak other languages. With this in mind, Fujitsu Laboratories has developed what it claims is "the world's first wearable, hands-free speech translation device."
Last year, Fujitsu developed a system in which stationary tablets with external microphones could identify different speakers, and translate their spoken language into another. Now, that technology has been shrunk down to a Wi-Fi-connected device that's worn by the worker, like a name tag.
It has two integrated omnidirectional microphones – one of them faces forward, to pick up the voice of the patient, while another faces up, to pick up the voice of the wearer. Onboard software automatically detects which of the two people is speaking, and a cloud-based server translates their words to synthesized speech in an output language of the user's choice.
This means, for instance, that before a conversation started, the user could stipulate that their own speech be translated to English, and that the patient's speech be translated to Japanese.
Additionally, the device is claimed to be very good at filtering out background noise. According to the company, the technology has "achieved a speech detection accuracy of 95 percent in an environment with comparable noise levels to an examination room in a large hospital (about 60 decibels of noise) at a natural distance for a face-to-face conversation between a healthcare provider and a patient of about 80 cm [2.6 ft]."
Plans call for clinical trials to begin in November, at the University of Tokyo Hospital and other healthcare institutions across Japan. Although the system currently only translates between Japanese, Mandarin and English, there are plans to expand the number of supported languages.