The advent of Internet Technology has brought human beings into a new environment in which users can get a great variety of information through computers. However, the interface between humans and machines is still cumbersome. We, the Spoken Language Communication Group, have set ourselves a goal of creating a technology that takes advantages of paralinguistic information such as speech and gestures, to make possible natural communication no matter who or where the speakers are, when and how they are used, or in which language they are communicating.

• Multilingual Speech Technology
  There are some factors that make accurate recognition difficult in a speech environment, such as various language characteristics, individuality of a person, different speaking-styles, and the acoustic conditions. Spoken Language Communication Group is working on developing robust multilingual speech recognition technologies, expressive multilingual speech synthesis technologies, and the acoustic signal processing technologies that control acoustic noises.
  
  
• Non-verbal information Technology
  Natural Language communication is possible not only by utilizing verbal information, but also by taking advantage of non-verbal information, such as speech, gestures, and situation. Aiming to develop a technology that can communicate naturally with Network terminals, we conduct research on analyzing paralinguistic and non-linguistic information, and on application technology.
  
  
• Spoken Language Technology
  To enable natural "human-to-human" or "human-to machine" communication, it is imperative that we develop a dialog system that processes a multimodal-dialog interface by appropriately controlling the timing of utterances, while supplying verbal and paralinguistic information such as intonation, facial expressions and gestures. We develop the technology for controlling utterance timing and construct a dialog model for appropriate responses, by utilizing paralinguistic information.
  
  
• To Attain an Intelligent, Next-generation Living Environment that is Kind to Everyone
  We conduct R&D on the universal platform technology that will enable a "Watch & Care" function in our society. The "Watch" performs to observe each individual from children to the elderly, and the "Care" carries out support for communication activities, outdoor activities, and lifelong learning, aiming to establish and intelligent next-generation living environment that is kind to everyone.
  
  
• Non-verbal Interaction
  We conduct a research on how a system offers an appropriate service in accordance with the non-verbal information obtained from glance, pointing, and physical movement of people. The system potentially provides the information service based on the user's interest, desire, and intention.