Department of Media Science

Speech and Sound Signal Professing and Language Information Processing

Kazuya Takeda

Professor, Department of Media Science

He has worked out researches on speech and sound processing and language processing such as sound signal processing ( reducing noises, symbolizing sound, and space sound processing ), speech information　processing ( recognizing speech and sensitizing speech ), and language information processing ( language models and dialogue systems ).

The speech is the closest method of communicating information to the human beings and the language is the most fundamental form of expressing their intelligence.

Therefore, sophisticating speech and language processing is essential to organize and develop information systems as social information infrastructures.

He has targeted to carry out high technology which is intended for this complicated real environment. He has researched multi channel signal processing methods, which is new and enables to be intended for whole human behaviors.

Searching for Possibilities of "Speech"

Norihide Kitaoka

Associate Professor, Department of Media Science

He has conducted researches on processing “Speech” which human beings utilize in the real environment by means of computer science such as sound recognition, multimodal dialogue systems, sound interface, and information search.

Speech has two rules; One is an interaction side to communicate with others and the other a work processing side to record and utilize knowledge. He would like to contribute to deep understanding of speech and image information by languages as well as to pursue computer linguistics in the real environment by dealing with　speech and image information and language information.

Modeling of Information Included in Human Voices and Actions

Chiyomi Miyajima

Assistant Professor, Department of Media Science

She has conducted researches on automatic recognition of individuality and intention included in human voices and actions.

She has statistically modeled information obtained from sound signals and moving graphics of movement of face, lips and both hands, which has helped develop recognition of individuals, speech, and sign languages.

She has promoted researches on recognition of individualities and intentions included in human behavioral signals such as operating accelerators, brakes and steering wheels while driving. She has hoped to apply this modeling of various human behavioral signals to systems to support safe driving and the auditorily handicapped.