• Speech Sciences
• /
• v.13 no.4
• /
• pp.43-50
• /
• 2006

The aim of this study was to identify pronunciation difficulties of Korean learners of English in their articulation of English sibilants /dg, g, z/. Forty-five syllables were produced five times by twelve college students. Test scores were measured from the score board made by FluSpeak, a speech training software program, which was designed for English pronunciation practice and improvement. Results show that 1) the subjects had lower scores in producing /g/ than /dg/ and /z/ from all positions, and 2) subjects had lower scores in inter-vocalic position than in pre-vocalic position and in post-vocalic position when they produced /dg/, /g/, and /z/. The results suggest that on the whole Korean learners have much difficulty in producing /g/, and they also have more auditory and articulatory problems in intervocalic than in the other positions when they produce these sibilants.

• Phonetics and Speech Sciences
• /
• v.4 no.2
• /
• pp.121-126
• /
• 2012

This study deals with how stuttering changes in its frequency in a situation where adult participants who stutter are exposed to one type of background noise, that is, multi-talker babble noise. Eight American English-speaking adults who stutter participated in this study. Each of the subjects read aloud sentences under each of three speaking conditions (i.e., typical solo reading (TSR), typical choral reading (TCR), and multi-talker babble noise reading (BNR)). Speech fluency was computed based on a percentage of syllables stuttered (%SS) and speaking rate was also assessed to examine if there was significant change in rates as a measure of vocal change under each of the speaking conditions. The study found that participants read more fluently both during BNR and during TCR than during TSR. The study also found that participants did not show significant changes in speaking rate across the three speaking conditions. Some discussion was provided in relation to the effect of multi-talker babble noise on the frequency of stuttering and its further speculation.

• Phonetics and Speech Sciences
• /
• v.4 no.2
• /
• pp.33-40
• /
• 2012

The purpose of this paper is to investigate the voice onset time (VOT) of the English voiceless stops [p, t, k] found in the Buckeye Corpus of Conversational Speech [1]. Three young female speakers were chosen for this study and their VOT values were semi-automatically extracted along with other factors. The factors used for the analysis were place of articulation, location in word, syllabic stress, content word or not, word frequency calculated from the corpus, and the speech rate expressed in syllables per second. Results showed that, for the three places of articulation of each speaker, all the factors had a statistically significant effect on the VOT values. This paper has significance in that the materials used for the analysis were from a corpus of spontaneous natural English speech.

• Phonetics and Speech Sciences
• /
• v.1 no.3
• /
• pp.109-115
• /
• 2009

Many isolated word recognition systems may generate non-similar words for recognition candidates because they use only acoustic information. In this paper, we investigate several techniques which can exclude non-similar words from N-best candidate words by applying Levenstein distance measure. At first, word distance method based on phone and syllable distances are considered. These methods use just Levenstein distance on phones or double Levenstein distance algorithm on syllables of candidates. Next, word similarity approaches are presented that they use characters' position information of word candidates. Each character's position is labeled to inserted, deleted, and correct position after alignment between source and target string. The word similarities are obtained from characters' positional probabilities which mean the frequency ratio of the same characters' observations on the position. From experimental results, we can find that the proposed methods are effective for removing non-similar words without loss of system performance from the N-best recognition candidates of the systems.

• Proceedings of the KIEE Conference
• /
• 1988.07a
• /
• pp.29-32
• /
• 1988

A new set of the units of demi-syllables is presented for Korean speech synthesis. The performance of the set of demi-syllable units is compared with that of the set of syllable units in the aspects of the quality of synthesized speech using each set of the units and the size of the computer memory which each set of units occupies. The set of demi-syllable units achieves comparable speech quality and occupies smaller memory size than the set of syllable units.

• MALSORI
• /
• no.63
• /
• pp.1-21
• /
• 2007

Locus equation scatterplots for [bilabial stop + vowel] syllables were obtained from 16 non-native speakers of English and Russian. The results indicated that both Russian speakers of English and English speakers of Russian exhibited modifications towards respective L2 norms in slopes and y-intercepts. All non-native locus equations generated exhibited linearity. Accordingly, the basic results reported in [17] were reverified by securing a larger subject base. More experienced speakers displayed better approximations to L2 norms than less experienced speakers, indicating the necessity of perception- and articulation-related learning for allophonic variations due to adjacent phonetic environments.

• Speech Sciences
• /
• v.10 no.4
• /
• pp.221-227
• /
• 2003

Diadochokinetic (DDK) speech task has been utilized as an evaluating tool for speakers with dysarthria for many years. This study attempted to differently diagnose multiple system atrophy (MSA) from idiopathic Parkinson's disease (PD) using patients' performance of DDK (i.e., alternate motion rate (AMR)). The subjects included 11 cases of pathologically confirmed MSA and 16 IPD patients who commonly presented with parkinsonian syndrome. The speech sample of each patient was analyzed acoustically using the MSPTM(Motor Speech Profile, a module of CSL). The results showed that the average DDK rate was significantly faster in the IPD than the MSA groups in all three syllables (i.e., /puh/, /tuh/. and /kuh/). We propose the average DDK rate variable as a core clinical trait in differentiating the two pathological conditions.

• Annual Conference on Human and Language Technology
• /
• 2016.10a
• /
• pp.9-13
• /
• 2016

Stack LSTM기반 의존 파싱은 전이 기반 파싱에서 스택과 버퍼의 내용을 Stack LSTM으로 인코딩하여 이들을 조합하여 파서 상태 벡터(parser state representation)를 유도해 낸후 다음 전이 액션을 결정하는 방식이다. Stack LSTM기반 의존 파싱에서는 버퍼 초기화를 위해 단어 표상 (word representation) 방식이 중요한데, 한국어와 같이 형태적으로 복잡한 언어 (morphologically rich language)의 경우에는 무수히 많은 단어가 파생될 수 있어 이들 언어에 대해 단어 임베딩 벡터를 직접적으로 얻는 방식에는 한계가 있다. 본 논문에서는 Stack LSTM 을 한국어 의존 파싱에 적용하기 위해 음절-태그과 형태소의 표상들을 결합 (hybrid)하여 단어 표상을 얻어내는 합성 방법을 제안한다. Sejong 테스트셋에서 실험 결과, 제안 단어 표상 방법은 음절-태그 및 형태소를 이용한 방법을 더욱 개선시켜 UAS 93.65% (Rigid평가셋에서는 90.44%)의 우수한 성능을 보여주었다.

• Annual Conference on Human and Language Technology
• /
• 2016.10a
• /
• pp.105-110
• /
• 2016

개체명 인식이란 문서 내에서 인명, 기관명, 지명, 시간, 날짜 등 고유한 의미를 가지는 개체명을 추출하여 그 종류를 결정하는 것을 말한다. 최근 개체명 인식 연구에서는 bidirectional LSTM CRFs가 가장 우수한 성능을 보여주고 있다. 하지만 LSTM 기반의 딥 러닝 모델은 입력이 되는 단어 표상에 의존적이기 때문에 입력이 되는 단어 표상을 확장하는 방법에 대한 연구가 많이 진행되어지고 있다. 본 논문에서는 한국어 개체명 인식을 위하여 bidirectional LSTM CRFs모델을 사용하고, 그 입력으로 사용되는 단어 표상을 확장하기 위해 사전 학습된 단어 임베딩 벡터, 품사 임베딩 벡터, 그리고 음절 기반에서 확장된 단어 임베딩 벡터를 사용한다. 음절 기반에서 단어 기반 임베딩 벡터로 확장하기 위하여 bidirectional LSTM을 이용하고, 그 입력으로 학습 데이터에서 추출한 개체명 분포를 이용하였다. 그 결과 사전 학습된 단어 임베딩 벡터만 사용한 것보다 4.93%의 성능 향상을 보였다.

• Annual Conference on Human and Language Technology
• /
• 2016.10a
• /
• pp.324-329
• /
• 2016

의미역 결정은 자연어 문장의 서술어와 그 서술어에 속하는 논항들 사이의 의미관계를 결정하는 것이다. 최근 의미역 결정 연구에는 의미역 말뭉치와 기계학습 알고리즘을 이용한 연구가 주를 이루고 있다. 본 논문에서는 순차적 레이블링 영역에서 좋은 성능을 보이고 있는 Bidirectional LSTM-CRFs 기반으로 음절의 의미역 태그 분포를 고려한 의미역 결정 모델을 제안한다. 제안한 음절의 의미역 태그 분포를 고려한 의미역 결정 모델은 분포가 고려되지 않은 모델에 비해 2.41%p 향상된 66.13%의 의미역 결정 성능을 보였다.