Yumiko Kato

Fundamental frequency pattern generator, method and program

According to this fundamental frequency generating method, a fundamental frequency pattern is set from a data base of a fundamental frequency pattern of each accent phrase standardized by the phoneme time length or the time length of the vowel and the vowel corresponding portion, and when the corresponding fundamental frequency pattern is not stored in the data base, the fundamental frequency pattern is generated by interpolating the interval between points serving as the references of the fundamental frequency pattern. With this method, a fundamental frequency pattern having higher naturalness than with conventional methods can be generated.

Speech synthesis method and speech synthesizer

A language processing portion (31) analyzes a text from a dialogue processing section (20) and transforms the text to information on pronunciation and accent. A prosody generation portion (32) generates an intonation pattern according to a control signal from the dialogue processing section (20). A waveform DB (34) stores prerecorded waveform data together with pitch mark data imparted thereto. A waveform cutting portion (33) cuts desired pitch waveforms from the waveform DB (34). A phase operation portion (35) removes phase fluctuation by standardizing phase spectra of the pitch waveforms cut by the waveform cutting portion (33), and afterwards imparts phase fluctuation by diffusing only high phase components randomly according to the control signal from the dialogue processing section (20). The thus-produced pitch waveforms are placed at desired intervals and superimposed.

Variable-speech-rate audiometry for hearing aid evaluation.

A new hearing aid evaluation method using variable-speech-rate audiometry (VSRA) was developed. VSRA was newly created based on the Japanese speech audiometry authorized by the Japan Audiological Society. The ordinary speech audiometry can not reveal a temporal factor in word discrimination ability of the hearing impaired. Since, with VSRA, we can compare several performance-intensity curves obtained from different speech-rate speech audiometries, the impact on the auditory system of each patient by the fast or slow speech rate could be easily determined. Taking the temporal factor of the auditory systems into consideration by using VSRA, hearing aid evaluation was performed for a master hearing aid with three types of signal processing and fitting for 36 hearing impaired subjects. Then hearing aid evaluation was performed using VSRA for a newly developed portable multi-function digital hearing aid with two types of signal processing and analog hearing aids which had been used by hearing-impaired patients. As a result, VSRA was useful for hearing aid evaluation, in particular, for cases when ordinary normal speech rate audiometry does not provide a significant difference in word discrimination scores. In addition, using VSRA revealed that amplitude compression is more effective for improvement of word discrimination than linear amplification.

Preliminary Study of Mobile Device-Based Speech Enhancement System Using Lip-Reading

This paper describes our preliminary study towards a new type of speech enhancement system. To avoid using odd-looking electrolarynx, we used lip-reading function. Our final image is to use a smart phone with camera and audio output to be able to convert the lip motion to speech output. We tested MLP, CNN, and MobileNets image recognition methods. 3k image datasets for training and testing were recorded from five persons. The preliminary experiment indicated that the MobileNets is the most adequate algorithm for smart phone apps. in terms of the recognition accuracy and the calculation cost.

Development of Hands-free Speech Enhancement System for Both EL-users and Esophageal Speech Users

A hands-free speech enhancement system for laryngectomies is proposed to improve the usability and the speech quality. The system has a small and right weight transducer, photo-reflector based lip movement level sensors, and a microphone. The sensor outputs are used for generating the transducer turn on/off signal. A simple performance test was conducted if the system can determine the utterance segments accurately, and the result showed good utterance detection performance. Also, we tested with five more subjects and confirmed that the proposed method was well performed even under the speaker-independent condition.

Electrode layout of wristband EMG measuring device for hand gesture recognition

In this study, we discuss a layout of electrodes on a wrist for EMG measurements to recognize hand gestures. Experimental results show that the proposed layout uses only half of the electrodes, but suffers only about 3% drop in recognition rate compared with the conventional layout. Furthermore, we developed a television operating system based on hand gestures, using a wristband-type EMG device with the proposed electrode layout.

Fixation distance estimation using vergence eye movement for automatic focusing glasses

In this study, we propose a method for estimating fixation distance on the basis of measurements of vergence eye movements. The aim of this approach is to control the lens focus of automatic focusing glasses. To reduce user effort at the time of calibration, the calibration was performed at infinite distance gazing, and the parameters were determined from the premeasured pupillary distance at infinity and iris diameters. To clarify the effectiveness of the proposed method, we conducted evaluation experiments using prototype glasses. The results showed that even participants requiring myopic correction could perform accurate motion vergence movements. Fixation distance estimation showed that, with the eye calibrated at infinite distance gazing, shorter distances could be estimated with an average accuracy exceeding 90%.