Spoken Language Research & Development
We mainly provide solutions utilizing our SpeeCAN® speech synthesis engine to manufacturers of disaster prevention administrative radio systems. We also provide speech synthesis and language processing technologies to manufacturers developing car navigation systems, TV games, and other products. In addition, we have been assisting national and foreign research institutes or companies to conduct part of their R&D work in speech recognition and speech synthesis.
Currently, our speech synthesis technology is shifting from the conventional corpus-based approach that utilizes large amounts of speech data to a deep learning (DNN (Note 2)) based method. We are also engaged in research and development of speech synthesis not only for Japanese, but also for English, Chinese, Korean, Indonesian, Vietnamese, and other languages.
(Note 2). DNN is an abbreviation for Deep Neural Network, a method of pattern learning that mimics the mechanism of human neurons.