Manabu Fukushima

Autonomous optimal location of articulation materials using robotics

We developed an application that determines optimal places of a robot equipped with articulation materials to improve sound environment. Lots of acoustic devices that incorporated architectural acoustics technology have been developed (e.g., articulation material, sound-absorbing material, etc.). In many cases, those materials are placed at positions that are effective by professional people. Our purpose is to make an autonomous robot to decide the materials position instead of professionals. We chose “Aural Sonic,” the articulation materials to be mounted on the robot. “Aural Sonic” is one of the acoustic devices that is expected effects to improve the hearing. This robot can change the position and angle little by little, to measure the impulse response each time. The application extracts the acoustic features corresponding to the speech clarity from the measured impulse response and calculates the evaluation value. The robot, which gives the position with high evaluation value, could be used for the a...

Autonomous mobile robot to improve sound environment for speech conversation

This study pursues to improve sound environment for speech conversation by moving an articulation panel by a robot. We tried to move the articulation panel to behind of the listeners by a small robot to keep fixed distance from the listener. An optimal location of the articulation panel depends on the sound sources such as speech conversation or music appreciation. It is usually decided based on the sense of users and experts. If we use a room for the purpose of speech conversation, the articulation panel is placed behind the listeners. However, moving the articulation panel optimally is almost impossible in terms of economical and physical conditions. In recent years, robot technology for transportation is in progress widely. We attempt resolving these problem by robotics. The articulation panel is framed by wood to have absorption and reflection surfaces, and the robot has omni-wheels for ultra-pivot turn performance. It can turn the surface on the sound source by the rotation of the robot. The robot ha...

Improvement of the accuracy of attenuation constant estimation using the cross spectral technique

This paper proposes a method to improve the accuracy of the attenuation constant estimate obtained using the cross-spectral technique. In the cross-spectral technique, the envelope of the estimated impulse response is deformed due to the use of a time window. As a result, the estimated impulse response decays more rapidly than the real impulse response does and the attenuation constant obtained by the estimated impulse response becomes larger than the real value. This paper explains first how the attenuation constant changes in the process of impulse response estimation, next, the proposed method for improving the estimation accuracy.

Deformation of impulse response estimates by the time window in the cross‐spectral technique

The estimation of the impulse response of a transfer system by the cross‐spectral technique is very popular and most FFT analyzers are equipped with that function. However, the deformation of the impulse response in the process of estimation is not yet known. The deformation of the impulse response is caused by the use of the time window in the FFT as the DFT of a windowed waveform is the convolution of the spectrum of the waveform and the spectrum of the time window. This paper first describes the theoretical analysis on the deformation. Next, the result is verified by the computer simulation. The estimation errors are also investigated for various types of time windows. According to the investigation, the deformation is significant if the duration of the impulse response (the reverberation time) is longer than 1/4 of the window length. Finally, the use of the rectangular window in order to not deform the impulse response by shortening the length of window for the source signal is described.

Computer assisted learning within the Smalltalk-80 environment

The authors discuss the potential of Smalltalk-80 for a teaching and learning environment. A laboratory-based introductory computer systems and programming course has been taught for more than three years within the Smalltalk-80 environment at the Department of Computer Science, Chiba Institute of Technology. The laboratory equipment consists of 45 networked workstations, in which Smalltalk-80 is implemented under the UNIX operating system. Courseware for the introductory course is embedded within the Smalltalk-80 environment for use by freshmen. Within the environment, students have been able to learn the course effectively based on a trial-and-error approach keeping their learning pace.<<ETX>>