We have been developing the speech visualization system as a speech training aid for the hearing impaired. This research has been carried out to build up a recognition system by which it is possible to represent a phoneme sequence on a display screen together with the visualized speech. This kind of system requires to recognize 20-50 isolated words which are spoken by anyone. Moreover, it is desirable that the vocabulary (20-50 words) of the dictionary can freely and easily be constructed or changed. Although the system which has phoneme standard patterns independently of the word dictionary is reasonable to achieve the aim, generally it will not always show a high recognition rate to a large vocabulary. In order to realize the high performance system which satisfies the above requirement, new speech parameters which are mutually complementary for the recognition have been adopted. A local distance between an input frame and each of 32 phoneme categories has been defined as a weighted sum of distances of the parameters because their dimensions are not the same. The weights have been decided to maximize phoneme recognition rates and tested for the word recognition. As a result, the addition of parameters was very effective and the recognition rates by the proposed parameters were much higher than by usually used one. When 50 words uttered by 30 males from two different databases were tested, the recognition rates were 97.2% and 98.3% in Bayes-decision-rule based distance.

View full abstract

Use of high grade damping materials is one of the effective methods to reduce vibration and noise, but the conventional damping materials have two weighty drawbacks, one is that it takes much labor to install, and the other is that it exhibits high damping performance at limited frequencies and temperatures. To overcome these drawbacks, we have developed a new type damping material Damping Material Applying Rubber Compounding Magnetic Powder (DRM). By compounding magnetic powder, DRM has not only a good workability but also enhanced damping abilities. As the result of experiments, the following properties are revealed.(1) The damping performance of DRM has a peak value on the magnetic attractive force.(2) The damping performance of DRM increases with a decrease in the ferrite content.(3) The damping performance of DRM increases with an increase in the vibratory amplitude while the conventional constraining type damping material has very little dependency on the vibratory amplitude. These properties are produced by the frictional loss generated between DRM and the vibrating body.

View full abstract

We propose a new data smoothing method for near-field acoustic holography (NAH) with the aid of the data extrapolation. This method can reduce the error energy which comes from the finiteness of the measurement aperture. We illustrate our calculational results by using of the computer simulations which apply to the numerically generated data by Fast Fourier Transform (FFT). The simulations show that the error energy leads to zero after 30 times iteration of extrapolation algorithm. Our calculations show that our data smoothing method is applicable for the quantitative analysis of the vibration energy in the NAH signal processing.

View full abstract

This paper is concerned with movements and stability of man gazing at a moving picture in a darkroom. These experiments studied to clarify the relation between visual information and postures. A worker may indirectly be exposed vibration when the objects around himswing. This is not only a problem for workers but also when he operates a large plant, it is a severe problem for the system if his posture swings and makes a mistake. The barpicture a subject gazes at was made with a computer in this experiments. In the first case, the center of gravity of a subject was measured when the bar swings right and left in front of him and the equilibrium line deviates gradually right or left. In the second case, the effects of gazing for a long time on subject's posture were measured. As results, our center of gravity deviates in the same direction as the equilibrium line deviates. And almost all subjects tune to the swinging bar in 5 min after they start their examinations, that is, this time may be the limit we can control our posture when we are gazing at a swinging bar in a darkroom.

View full abstract

This paper describes the results of combining several hybrid transducer ultrasonic motors together in parallel on a single central shaft in order to obtain a large output torque. The authors propose a practical connection mechanism using the hollow cylindrical motors. The rotor of each motor is connected together via a rod which penetrates through the center of each motor, and thus, the rod collects the output torque from each of the motors; the accumulated torque appears at the end of the rod. In a theoretical study, an equivalent electrical circuit model is introduced to represent the torque accumulation mechanics and it indicates that the maximum torque of the combined system is equal to the summation of the maximum torque obtained from each motor. Two combinations of the system composed of two motors are tested in an experimental study. The measured maximum torque of each system is equal to the summation of the maximum torque obtained from each motor as shown in the theoretical investigation.

View full abstract

An efficient segment-matching method is presented for blood-flow velocity estimation using pulsed ultrasound RF echo-signals. In contrast to other motion-tracking methods, this approach estimates the blood flow based on the well-known Eulerian standpoint rather than the Lagrangian standpoint used in most of time-domain correlation techniques in medical ultrasound. The time shift corresponding to the displacement the local scatterers (blood cells) moved during the pulse repetition period is assessed by comparing the patterns of each pair of sampled segments within a predefined search area. All pairs of sampled segments used for comparison are selected to be symmetrical about the observation point inside the pair of consecutive echo-signals. The error analysis via computer simulation demonstrates that the Eulerian segment-matching method proposed here is more suitable and more reliable for the case of blood-flow velocity estimation relative to the Lagrangian method. The maximum detectable velocity which is limited by the size of the search area in time-domain techniques can be improved twice as much as that in conventional tracking methods, and the percentage of false velocity estimates as a function of velocity distribution of multiple scatterers is also lower than that in the Lagrangian tracking method.

View full abstract

For the purpose of attaining higher accuracy of the 3D velocity quantitative CT image of the biological tissue based on the inverse scattering analytical method, introduction was made with such improved methods as of quasi-3D image reconstruction, correction of the 2nd order bias of the Rytov approximation based on the data collection on the aperture of a cylinder, together with its performance evaluation. As a result of the value test by means of the simulation for a sphere or cylindrical object, it is verified that a sound velocity quantitative CT image with by far higher accuracy and higher quality than the conventional one can be brought to realization under the condition of the sound velocity ratio and observation distance satisfying requirement for practicality.

View full abstract

Marker combination effects were found in the discrimination of auditory durations using short target parts of tones marked by rising and/or falling level changes. The obtained JNDs were the smallest for durations marked by two rising changes (rise-rise type) or for those of the rise-fall type; moderate for those of the fall-rise type; and the largest for those of the fall-fall type.
Although the variations in the experimental conditions were limited, they certainly provided direct evidence for (1) the predominance of rising markers to falling markers, and for (2) the predominance of rise-tofall durations to the reverse case, fall-to-rise durations, in the discrimination of auditory durations.

View full abstract