• The Journal of the Acoustical Society of Korea
• /
• 제18권1E호
• /
• pp.27-31
• /
• 1999

Recent linguistic researches have brought into focus the relations between prosodic structures and syntactic, semantic or phonological structures. Most of them prove that prosodic information is available for understanding syntactic, semantic and discourse structures. But this result has not been integrated yet into recent Korean speech recognition or understanding systems. This study, as a part of integrating prosodic information into the speech recognition system, proposes an automatic detection technique of Korean accentual phrase boundaries by using one-stage DP, and the normalized pitch pattern. For making the normalized pitch pattern, this study proposes a method of modified normalization for Korean spoken language. For the experiment, this study employs 192 sentential speech data of 12 men's voice spoken in standard Korean, in which 720 accentual phrases are included, and 74.4% of the accentual phrase boundaries are correctly detected while 14.7% are the false detection rate.

• The Journal of the Acoustical Society of Korea
• /
• 제23권1E호
• /
• pp.10-15
• /
• 2004

In performing speech conversion from a source speaker to a target speaker, it is important that the pitch contour of the source speakers utterance be converted into that of the target speaker, because pitch contour of a speech utterance plays an important role in expressing speaker's individuality and meaning of the utterance. This paper describes statistical algorithms of pitch contour conversion for Korean language. Pitch contour conversions are investigated at two 1 evels of prosodic phrases: intonational phrase and accentual phrase. The basic algorithm is a Gaussian normalization [7] in intonational phrase. The first presented algorithm is combined with a declination-line of pitch contour in an intonational phrase. The second one is Gaussian normalization within accentual phrases to compensate for local pitch variations. Experimental results show that the algorithm of Gaussian normalization within accentual phrases is significantly more accurate than the other two algorithms in intonational phrase.

• 음성과학
• /
• 제14권3호
• /
• pp.7-19
• /
• 2007

The present study investigated some phonetic attributes which distinguish two Korean stop types $^-aspirated$ and $lax^-$ in a prosodic position which has previously received little attention, namely medial in an accentual phrase. The intonational pattern across syllables which are initial in an accentual phrase (Jun, 1993) is said to depend on the type of stop (aspirated or lax), while that of syllables which are medial in an accentual phrase are not. In Experiment 1, nine native Korean speakers read sentences with a controlled prosodic pattern in which aspirated or lax stops occurred in accentual phrase-medial position. Acoustic analysis revealed significant differences between aspirated and lax stops in closure duration, voice-onset time, and fundamental frequency (F0) values for post-stop vowels. The results indicate that a wider range of acoustic cues distinguish aspirated and lax Korean stops than previously demonstrated. Phonetic and phonological models of consonant-tone interactions for Korean will need to be revised to account for these results.

• 한국음향학회지
• /
• 제18권8호
• /
• pp.9-16
• /
• 1999

일반적으로 사람은 말을 할 때 어절들은 몇몇의 구로 그룹핑하여 발음함으로써 발화한다. 이것은 듣는 사람으로 하여금 발화의 의미와 의도를 잘 파악하도록 도와준다. 특히, 이러한 목적으로 발화자는 무의식적으로 운율정보(억양, 장단, 리듬 등)를 적절히 사용하게 된다. 본 논문에서는 발화된 문장에서 운율경계를 인식의 단위로 하는 음성인식방법에 대하여 제안한다. 즉, 발화된 문장을 운율구단위로 나누는 방법을 제안하고 나누어진 단위에 따라 연속음 인식실험을 수행하였다. 인식실험결과 연속음인식 시간의 감소를 관찰할 수 있었으며, 물론 음성인식률도 20-10%정도 증가하였다.

• 한국음향학회지
• /
• 제28권2호
• /
• pp.155-163
• /
• 2009

Text-to-speech 시스템에서 입력 텍스트로부터 운율 정보를 생성하기 위해서는 운율구 경계, 음소 지속시간, 기본주파수 포락선 설정의 3가지 기본적인 모듈이 필요하다. Break 인덱스 (BI; Break Index)는 합성기에서 운율구의 경계를 나타내고, 자연스러운 합성음을 생성하기 위해서는 BI를 정확히 예측하여야 한다. 그러나 BI는 문장의 의미나 화자의 읽기 습관(reading style)에 따라 임의적으로 결정되는 경우가 많아 정확한 예측이 매우 어렵다. 특히 일본어 합성기에서는 악센트 구 경계 (APB; Accentual Phrase Boundary)와 major phrase 경계 (MPB; Major Phrase Boundary)의 정확한 예측이 어렵다. 따라서 본 논문에서는 APB와 MPB 예측 오류를 보완할 수 있는 방법을 제안한다. BI를 고정 break (FB; Fixed Break)와 가변 break (VB; Variable Break)로 분류하여 합성단위 선택을 수행한다. 일반적으로 BI는 한번 생성되면 변하지 않는다. 따라서 BI가 잘못 생성된 경우 최적의 합성음을 생성할 수 없게 되는데, VB는 생성된 BI와 그것과 유사한 BI를 함께 이용하여 합성단위 선택을 수행함으로써 합성음의 BI가 생성된 BI와 다를 수 있는 것을 의미한다. APB와 MPB에 해당하는 BI에 대하여 VB인지 FB인지 CART(Classification and Regression Tree)를 이용하여 예측하고, VB인 경우 기본 주파수와 음소 지속시간에 대해 다중 운율 모델을 생성하여 합성단위 선택을 수행하였다. MOS 테스트 결과 원음이 4.99, 제안한 방법을 4.25, 기존의 방법은 4.01로 합성음의 자연성을 향상시킬 수 있었다.

• 대한음성학회:학술대회논문집
• /
• 대한음성학회 1996년도 10월 학술대회지
• /
• pp.323-332
• /
• 1996

The current speech technology has been aiming to acquire much clearer and more natural synthetic speech sound. The naturalness can be developed by an adequate phrasing of target sentence, of course, which seems to be strongly related to both syntactic and phonetic aspect simultaneously. The present study aims to describe, at one aspect, the relatedness between syntactic structure and prosodic phrasing through dialogue speech, and at the other, to establish a suitable phrasing pattern with respect to the purpose of acquiring more natural synthetic sound. The prosodic phrase, here, means a prosodic unit which can be clearly identified as having an evident break boundary at its final position in a sentence in the sense of both perceptual and acoustical viewpoint. The end of each prosodic phrase is, accordingly, marked as the point of major boundary in a sentence.

• 음성과학
• /
• 제11권1호
• /
• pp.31-42
• /
• 2004

This paper presents methods using Gaussian normalization for converting pitch contours based on prosodic phrases along with experimental tests on the Korean database of 16 declarative sentences and the first sentences of the story of 'The Three Little Pigs'. We propose a new conversion method using Gaussian normalization to the pitch deviation of pitch contour subtracted by partial declination lines: by using partial declination lines for each accentual phrase of pitch contour, we avoid the problem that a Gaussian normalization using average values and standard deviations of intonational phrase tends to lose individual local variability and thus cannot modify individual characteristics of pitch contour from a source speaker to a target speaker. From the results of the experiments, we show that this slanted Gaussian normalization using these declination lines subtracted from pitch contour of accentual phrases can modify pitch contour more accurately than other methods using Gaussian normalization.

• 대한음성학회지:말소리
• /
• 제39호
• /
• pp.15-24
• /
• 2000

This study aims to reexamine Korean Lenis Stop Voicing (henceforth, LSV) and to specify its phonetic conditions in phonetic terms. LSV optionally occurs within certain prosodic domains. They are called 'Malthomak'(Lee, 1996),'phonological phrase'(Kang, 1992), or 'accentual phrase'(Jun, 1993). On the basis of Jun's phrasing, this study focuses on the more specific phonetic conditions of LSV in the accentual phrase medial position, sub-classifying voicing as complete and partial. The results shows that whether the stops become completely voiced or partially voiced was determined by the various phonetic environments, such as adjacent segments and following intonational phrase boundaries. It is shown that the conditions of LSV should be described in terms of more detailed phonetic environments and that they could be used in predicting the class of voicing.

• 대한음성학회지:말소리
• /
• 제46호
• /
• pp.127-143
• /
• 2003

This paper presents segmental labeling conventions proposed by SiTEC (Speech Information Technology Engineering Center) 2002 and proposes a new directions of a revision for a simpler version. The paper also reviews one of the prosody labelling conventions for Korean, K-ToBI convention(ver. 3.1) and proposes a couple of modifications and suggestions.

• 음성과학
• /
• 제7권1호
• /
• pp.143-169
• /
• 2000

This chapter presents an overview of Korean intonational structure and proposes a revised version of K -ToBI (Korean TOnes and Break Indices), a prosodic transcription convention for Seoul Korean. In the new version of K-ToBI, a tone tier is separated into two tiers: a phonological tone tier and a phonetic tone tier. A phonological tone tier labels tones marking the prosodic structure of an utterance, and a phonetic tone tier labels individual tones of an AP and an IP conforming to the surface pitch contour. Labelling surface tonal patterns will provide us data to test the underlying tonal patterns and to build phonetic implementation rules.