Research Areas

The following fields are research at the media information laboratory.
Media Search
We are developing very quick multimedia search and recognition methods,
which will enable us to instantaneously retrieve desired audio/video
segments from several years' worth of unlabelled audio/video archives.
[Recognition Research Group ]

Media Interaction
Our research focuses on measurement of human behavior
in using various kinds of information media
and communication supporting systems utilizing these measurement.
[Recognition Research Group ]

Blind Source Separation
Sound source separation is a technique for separating a target sound
from mixtures of signals. Our research efforts have been focused on
developing a separation method based on Independent Component
Analysis (ICA).
[Signal Processing Research Group]

Speech Recognition
Aiming at accurate and natural human-computer communication, we are
developing a next-generation open-vocabulary speech recognition
system that can be used robustly in various real-world communication
scenarios.
[Signal Processing Research Group]

Dereverberation
Dereverberation is referred to as the technique to remove the effects of
reverberation from reverberant speech. Our research goal is not only
recovering highly intelligible speech, but also developing an efficient
pre-processing technique for automatic speech recognition in a
reverberant environment.
[Signal Processing Research Group]

Augmented-Reality Communication
Our goal is to research and develop a room-sharing video system "t-Room" that allows people to simultaneously experience distant space and remote time.
[Communication Environment Research Group]

Ambient Intelligence
Our research focuses on exploring activation mechanism for communication through appropriately effecting changes in human emotions, feelings, and intelligence.
[Communication Environment Research Group]