There are two types of human psychological impression caused by noise stimuli. One is an instantaneous and temporal impression following randomly fluctuating noise level (psychological impression in the short time scale). The other is an overall psychological impression which occurs after the person is exposed to the noise stimuli for a certain time interval (psychological impression in the long time scale). In the previous papers, with an unavoidable fuzziness attached to subjective judgment of human beings, a method for evaluation of the psychological impression in the short time scale has been studied. This paper is a further evolution of the considerations already published. That is, we focus on estimating and/or predicting the psychological impression in the long time scale by using the psychological impression in the short time scale. Concretely, we conduct an indoor psychological experiment to establish a psychological impression to in the short time scale and long time scale which are divided into seven categories. We then propose an estimation and/or prediction method of the psychological impression in the long time scale using the well known linear regression equations. Furthermore, a method for estimating and/or predicting the psychological impression in the long time scale is also considered by use of the level statistics of the noise stimuli.

View full abstract

A classification method for 3D body forms was proposed by the present authors using the Free Form Deformation technique. Whereas this method is time consuming and unsuitable for processing large amount of data. In the present study, the 3D foot forms of 63 adult female subjects were analyzed by this method, with four scales of variation being extracted by multidimensional scaling. Estimation equations were calculated to estimate the scores of the four scales by the stepwise multiple regression using bootstrap resampling techniques. In total, 11 variables, which are easily measured by hand, were adopted for the equations. Using the estimation equations, the distribution of 3D foot characteristics was investigated for 204 subjects. The four scales for 3D foot forms were related to the 1) higher versus lower instep contrast, 2) anteroposterior proportion, 3) medial inclination of the heel and ankle, and 4) direction of the toes. Multiple correlation coefficients between observed and estimated scores of the four scales were 0.85-0.92. Scale 1 was related to foot length size, and scale 2 to foot girth size. Very small feet (foot length <222mm) tended to be high arched, and very large feet (foot length>248mm) tended to be flat. Very narrow feet (foot girth size B or C) tended to have long toes and short ankle depth, and very wide feet (foot girth size EEEE or F) tended to have short toes and long ankle depth. Subjects at the distributional extremes having either small or large feet accounted for about 20% of the sample. Shoe comfort for these people would be improved by introducing their shape characteristics into the design of shoe lasts.

View full abstract

For people with hearing disabilities, a device was developed which can translate speeches into Japanese sentences with only a small delay. In order to lessen the fatigue of readers, the readability of texts on large screens was studied. The size of the screen was about 1200mm wide and 780mm high. The resolution of large screen display was SVGA (800*600 pixels). Six subjects read Japanese sentences from the screen, the variables being character-viewing angles and viewing distances. The average reading speeds and the margins of CFF before and after reading were measured. Also, the fatigue that the subjects felt and the readability of each sentence were surveyed by questionnaires. As a result, the most readable character-viewing angle was estimated about 0.011rad. at 2m viewing distance. They were about 0.008rad. and 0.0065rad. at 4m and 8m viewing distance, respectively. That is, the most readable character-viewing angle decreases gradually as the viewing distance increases. However, other factors, such as the pixel-matrix size of a character, the number of characters per line and the number of lines per page were different on each experiment. And so, there is a possibility that they might influence the results to some extent.

View full abstract