|Fujitsu Introduces Developments to Speech Synthesis Technology|
Fujitsu Laboratories Ltd. has developed a speech synthesis technology which produces high-quality synthetic voices, and transmit information in a tone that is appropriate for any scenario.
At present, speech synthesis devices can read texts, but deliver it in monotone. Thus, giving messages the proper tone helps recipients to understand and respond accordingly.
The more sophisticated technology from Fujitsu can create voices in 1/30th of the time that is previously required. Likewise, messages can be customized according to need. This enables texts to be translated to alarming messages in times of emergencies and to subtle tones when explanations are needed to be passed on.
Originally, speech synthesis has only been used to convert text into audio messages. But with an ever-changing technology landscape, this feature has been integrated into more practical usage. This is highly evident in car navigation systems, interactive voice responses and centralized audio announcements in offices or schools.
Though it has existed for quite a while now, conventional speech synthesis output disregards pitch and speed. This can create confusion or disparity between the actual response and the needed response from the recipients. For voice services, there is also a constant need to create new voices that will be used in operations.
Unlike typical synthesis methods wherein pre-recorded speech is recorded in waveforms, Fujitsu’s speech synthesis devices incorporate intonation, diction and voice quality which can be adjusted to specifications. This way, the messages become more authentic and those who will hear it can respond accordingly.
On the other hand, Fujitsu’s take on speech synthesis rely on an algorithm that banks on machine learning. This lets the device to extract the text efficiently and incorporate related parameters quickly. A shorter creation time translates into a more efficient message exchange. Of course, adding the proper tone and emotion to the correspondence ensures that there will not be any discrepancies in interpretation.
Aside from providing a more transparent information exchange in the corporate setting, Fujitsu’s advanced speech synthesis technology can be vital during communal announcements for natural disasters. Sending the right message enables the concerned parties to prepare accordingly, especially when the perceived effects are severe.
Also, speech synthesis can be used to record the speech of those who are about to lose their voice due to an illness. With the pre-recorded pieces of information, the person can still continue to communicate despite his condition.
Despite having a more advanced speech synthesis system, Fujitsu Laboratories is still making further improvements. Barring any delays, they can usher in this technology into practical usage during the 2014-2015 fiscal year.