荒井 隆行

To achieve ultra-realistic communications (URC), a three-dimensional (3D) sound field recording and reproducing system using spatial sound features is promising. It utilizes sound source positions and original source signals as features. On the other hand, sound source orientations (such as speaker's orientation) have been neglected, though they are necessary for high-performance URC. In this paper, we propose a 3D sound field recording and reproducing system that includes sound source orientation. For the recording part, we applied and evaluated a sound source orientation estimation method based on orientation-extended beamforming. This method requires transfer functions for all possible source positions and orientations, which should be either measured or calculated. Measured transfer functions have high accuracy because all characteristics, such as reverberations and diffractions, are taken into account. Calculated transfer functions are obtained for any environment without measurements. We performed experiments to evaluate the orientation estimation method using transfer functions obtained by both practical measurements and calculations using acoustic simulation based on wave theory. The experimental results show that our estimation method has sufficient capability using both measured and calculated transfer functions. For the reproducing part, we evaluated impulse responses versus sound source orientation. The evaluation results revealed significant differences in both objective and subjective scores. This proves that our system achieves higher performance than conventional ones that do not utilize orientation features. ©2010 IEEE.

Numerous research has investigated how first language influences the perception of foreign sounds. The present study focuses on the perception of voiceless English fricatives by Japanese listeners with advanced and intermediate level English proficiency, and compares their results with that of English native listeners. Listeners identified consonants embedded in /a _ a/ in quiet, multi-speaker babble and white noise (SNR=0 dB). Results revealed that intermediate level learners scored the lowest among all listener groups, and /th/-s/ confusions were unique to Japanese listeners. Confusions of /th/-/f/ were observed among all listener groups, which suggest that those phoneme confusions may be universal.

The current paper discusses two approaches to enhanced speech in reverberation/noise: machine signal processing and human speech production. We reviewed the speech enhancement techniques, including steady-state suppression and compared the modulation spectra of speech signals before and after processing. We also introduced the Lombard-like effect of speech in reverberation, and compared the characteristics of speech signals, including the modulation spectra between speech signals uttered in quiet and reverberation. We found that the enhanced speech signals have distinct characteristics that yield higher speech intelligibility.

Arai has developed several physical models of the human vocal tract for education and has reported that they are intuitive and helpful for students of acoustics and speech science. We first reviewed dynamic models, including the sliding three-tube (S3T) model and the flexible-tongue model. We then developed a head-shaped model with a sliding tongue, which has the advantages of both the S3T and flexible-tongue models. We also developed a computer-controlled version of the Umeda & Teranishi model, as the original model was hard to manipulate precisely by hand. These models are useful when teaching the dynamic aspects of speech.

We tested how a time-reversed signal effectively masks a target speech signal by reversing the time frame preceding the target frame to obtain a masking signal. We confirmed that this proposed masker, which is robust against speaker change and has advantages for implementation on a DSP chip, can effectively mask an input speech signal. In fact, a proposed masker was recently implemented with a DSP in our laboratory. We will further investigate the performance of the algorithm to produce the masking signal by increasing the number of stimuli and listeners. © 2010 The Acoustical Society of Japan.

Previous research has revealed that Japanese native speakers are highly likely to both perceive and produce epenthetic vowels between consonants. The goal of the present study is to investigate the influence of English learning backgrounds in the perception and production of consonant clusters by Japanese native speakers. In Experiment 1, a forced-choice AXB task to identify VC(u)CV is assigned to 17 highly fluent Japanese-English bilinguals and 22 Japanese monolinguals. Results show that monolinguals made significantly more errors than bilinguals. In Experiment 2, the influence of English proficiency on the production of consonant clusters, and the effect of consonant voicing on vowel epenthesis are investigated. The epenthetic vowels are acoustically analyzed and categorized into three degrees: full, partial and no epenthesis. The voicing combinations of the consonant clusters are C[+voice]-C[+voice], C[−voice]-C[+voice], and C[−voice]-C[−voice]. Results show that monolinguals inserted more epenthetic vowels than bilinguals, and that the influence of consonant voicing was stronger in monolinguals than bilinguals. Furthermore, monolinguals' epenthetic vowels between C[−voice]-C[+voice] and C[−voice]-C[−voice] tended to become devoiced than bilinguals. This result suggests a stronger L1 influence on monolinguals. The results of the two experiments thus suggest that the English proficiency influences the perception and production of consonant clusters.

While listeners are recognizing words from the connected speech stream, they are also parsing information from the intonational contour. This contour may contain cues to word boundaries, particularly if a language has boundary tones that occur at a large proportion of word onsets. We investigate how useful the pitch rise at the beginning of an accentual phrase (APR) would be as a potential word-boundary cue for Japanese listeners. A corpus study shows that it should allow listeners to locate approximately 40-60% of word onsets, while causing less than 1% false positives. We then present a word-spotting study which shows that Japanese listeners can, indeed, use accentual phrase boundary cues during segmentation. This work shows that the prosodic patterns that have been found in the production of Japanese also impact listeners' processing.

日本語発話における変調スペクトルとリズムを構成する単位の関係を調べた。変調スペクトルは,音声の単位の出現するタイミングと関係がある。英語の音節の方が日本語よりも長いにも関わらず,英語発話と日本語発話の変調スペクトルは類似している。「日本語MULTEXT」から得られた変調スペクトルのピークは,4〜5Hzであり,200〜250msの時間長に相当する。モーラの長さは短く,変調スペクトルのピークには対応しない。音節の長さは,ほとんどのものが短かったが,2モーラ以上から構成されている音節は200〜250msに分布していると推定された。連続する2モーラおよび2音節の長さは200〜250msであった。これらのことから,単独のモーラや音節ではなく,2モーラまたは2音節からなる脚が変調スペクトルのピークに対応していることが分かる。変調スペクトルと2モーラまたは2音節からなる脚の関係を示すことができた。2モーラまたは2音節の脚は,かなり一定の間隔で現れ,強さ曲線への影響が大きい。

Investigation on human speaker identification enables us to know the indexical cues to speakers, and it may consequently lead to the effective acoustical parameters that can be used for forensic speaker recognition. It is known that speaker individuality interacts with the phonological or linguistic information contained in speech signals. As proof, the accuracy of perceptual speaker identification (PSI) performances depends on what types of sounds are presented to the listeners. In a series of our previous experiments, we have been investigating the effective sounds for PSI, and the stimuli containing a nasal were found to be the ones. In this present study, we conducted another PSI experiment in order to examine the reproducibility of the nasal effectiveness, and to see the effects of the following vowels. Coronal nasals were shown to be effective despite the different speaker set or-the following vowels, and the stimuli containing a nasal were significantly better than those without it. In the second part of this paper, we introduce the results of the acoustical analysis of the stimuli. The contours of the energy transitions showed variations in shape among speakers for all three types of the analysis targets; nasals, stops, and fricatives, although the inter-speaker difference in the energy slopes for the consonant articulation was significant especially in nasal sounds. We also examined the effects of the sampling frequencies and the speech codecs, and found that the speaker-dependent shapes of these energy contours were maintained as long as the speech materials were uncompressed. The contours of the nasals appeared to be stable within a speaker, compared to other types of sounds. (c) 2008 Elsevier Ireland Ltd. All rights reserved.

This study investigates the characteristics of the two major dialects of Japanese: Osaka and Tokyo dialects. We recorded the utterances of the speakers of both dialects, and analysed the differences that appear in the accentuation of the words at the phonetic-acoustic level. The Japanese words that are phonologically identical in both dialects were used as the analysis target. The results showed that the pitch patterns contained the dialect-dependent features of Osaka Japanese. Furthermore, these patterns could not be fully mimicked by speakers of Tokyo Japanese. These results show that there is a phonetics-phonology gap in the dialectal differences, and that we may exploit this gap for forensic purposes.

In the speech-related field, physical models of the vocal tract are effective tools for education in acoustics. Arai's cylinder-type models are based on Chiba and Kajiyama's measurement of vocal-tract shapes. The models quickly and effectively demonstrate vowel production. In this study, we developed physical models with simplified shapes as educational tools to illustrate how vocal-tract shape accounts for differences among vowels. As a result, the five Japanese vowels were produced by tube-connected models, where several uniform tubes with different cross-sectional areas and lengths are connected as Pant's and Arai's three-tube models.

In a previous study, Arai implemented a sliding vocal-tract model based on Fant's three-tube model and demonstrated its usefulness for education in acoustics and speech science. The sliding vocal-tract model consists of a long outer cylinder and a short inner cylinder, which simulates tongue constriction in the vocal tract. This model can produce different vowels by sliding the inner cylinder and changing the degree of constriction. In this study, we investigated the model's coverage of vowels on the vowel space and explored its application for vowel production in the speech and hearing sciences.

The effect of steady-state suppression (SSS) in situations where listeners hear natural and electroacoustical sounds, was investigated. The study found out that how multiple sound source listeners hear, how loudspeaker gain, and how the input to the PA system affect the performance of SSS. The listening test on young people with normal hearing using natural and steady-state suppressed speech under simulated reverberant environments in which PA systems were virtually installed on computer, was done to verify the study. Results from these tests showed that SSS improved syllable identification in cases when listeners receive sounds either from a single source or from two sources. Increasing the loudspeaker gain changed the performance of SSS, and the effect of the mixture of electroacoustical sounds with sounds from a talker at the talker microphone is found to be negligible when the direct to reverberation ratio is high.

There are several factors that affect human speaker recognition. In this study, two experiments were conducted in order to see the effects that the stimulus contents and the familiarity to the speakers give to the perception of the speakers. The results showed that: a) stimuli including a nasal were effective for accurate speaker identification b) coronal nasals were more effective than the labial nasal, and c) the familiarity to the speakers gives a great influence on the performance. The tendencies a) and b) were observed both in familiar and unknown speaker identifications. The results of the acoustical analyses also showed that there were correspondences between the perception of the speaker identity and the cepstral distances among the speakers. The inter-speaker cepstral distances were greater in vowel intervals than in the consonant intervals especially, notably they were greater in nasals than in orals in the consonant intervals. © 2009 The Acoustical Society of Japan.