Current Research

Modeling of Human Information Processing Mechanism and its Application

OUTLINE
 Humans have an excellent information processing mechanism. We clarify such an information processing mechanism concerning the scientific search to discover the human mechanism. We expect that the mechanism will be applied to engineering.
 By imitating the human information processing function, we can discover new techniques that realize a learning machine. We can also obtain basic data for the design of a human friendly interface by considering human characteristics.
 By developing computer engineering, methods of measuring biological signal are changing. To improve presentation and recording methods, we can estimate more detailed functions.
 We have designed an original experimental system for biological signal measurement and tackled the modeling of the information processing process and its applications.
TOPICS
Examination of the role of ocular parallax caused by small eye movements as depth perception cues
Recently, attention has been focused on the motion parallax induced by eye movement as a role of eye movement. However, this parallax was confirmed, when large sight line shift, which was not observed in ordinary observation, occurred. Here, we examined the effect of parallax in the sight line shift in ordinary observation. We obtained results indicating that parallax is utilized as a depth cue by the measurement of eye movement. We conjecture that eyeball movement, for example, fixation nystagmus, functions as depth cue.
Computer vision system based on depth estimation by parallax caused by small rotation of camera
Using the above results, we tried to develop depth information presentation equipment as an engineering application by repeating the small rotation of the visual axes of a camera, as in fixation nystagmus. The visual camera axis is not fixed, but it shifts in small steps to measure the depth. The system integrates the depth results by individual small sight line shifts.
Extension of a model of induction field in vision
 We seek engineering applications of an"induction field in vision,"which is the fundamental characteristics based on experimental psychology results. A model for a binary image as a conventional model has represented this induction field. Therefore, we cannot currently apply the model for gradation or color images. So, we must measure the characteristics of an induction field in vision under various conditions and tackle the problem that expands the model under various conditions. We have pointed out that consideration of line width is also necessary when the model is constructed. At present, we are examining the effects of color on the construction of the induction field model.
FUTURE WORK
 Thus, this research field clarifies human information processing mechanism. We hope that these results are useful for the development and the creation of human friendly interfaces. Since measurement techniques continue to advance, we can obtain the data for constructing more detailed model.
 Until now, we have mainly researched vision. Humans have other important sensory mechanisms. We consider it an important mission to clarify the information processing mechanism or interaction processes of other senses.