Our goal is to understand the mechanism of information processing in the brain, specifically, motor control and visual perception. One major topic is the process of how the brain performs and acquires skilled motor movements. To approach this problem, we develop computational models based on robotics, control theory, and learning theory, together with running behavioral experiments. Another major topic is the vision system. We develop models for early and middle vision, based on mathematical, statistical and computational methods, to explain experimental findings. These models can also contribute for developing novel algorithm of image processing. We are also interested in the cooperation, coordination and interaction between sensory and motor systems because our sophisticated functions are emerged from their coordination.
ABCUnderstanding how the brain controls our body is central to solve a mystery of intelligence of human being, since most of intellectual activities and skills involve our body movements. Especially, the neural mechanism for generating motor command is quite important because the brain has to somehow coordinate hundreds of muscles at the same time. How does the brain compute to generate the coordinated muscle activities?
For example, consider a scenario where you make a simple reaching movement to grab a coffee cup. First your vision identifies the location of the coffee cup. This target location is represented in the eye centered coordinate systems. Meanwhile, you should identify the location of your hand by taking information from the peripheral sensory organs. Then, you compute the initial movement direction after the coordinate transformation from eye centered to hand coordinate systems. To produce a smooth movement, you also need to compensate the inertia of the arm. Like this, the process of making an even simple movement is very complicated. We are puzzled by the fact that the human brain is able to complete this computation in 200 ms. What is the neural mechanism in the brain under this process?
To answer this question, we develop a computational theory of motor control based on the control theory, robotics and learning theory. To examine this model, we measure human motor behavior. To understand the neural mechanism further, we also collaborate with neurophysiologists in other universities and laboratories and study characteristics of neural activities for motor control.
The visual processing has been central for human to survive. Human brain acquires this high performance system in the process of the development. Now, many engineers are trying to design artificial machines for this splendid function. One possible approach to make a breakthrough is to study how the brain processes visual information and to implement the computational process of the brain into computer algorithms. To tackle this problem, we are trying to answer how a simple neural model can achieve mathematical operations such as derivation and statistical inference, which are essential for visual image processing. We also examines the behaviors of these models by computer simulations together with running psychophysical experiments for testing the validity of the proposed models.
Artificial sensory channel: A tool for human assistance and human research
It is well known that, since human beings appeared in this world, they have succeeded to extend their ability by developing many new tools. Nowadays, engineers are trying to extend abilities of human by using the technology of virtual reality where systems present multi-modal information to help human perceive what they had difficulty to sense.
Especially, we are developing a technique to present auditory signals to lead us to perceive what we cannot sense by ourselves. For example, we are not conscious about our posture because we can’t see it directly. If we “audible-ize” this posture, it might help us to perceive our posture and acquire motor skills efficiently. Sensory enhancement through such artificial sensory channel could provide a new tool for assist the humans, and at the same time, it could be an experimental tool for examining how such additional sensory information modulates human functions.