• The Journal of the Acoustical Society of Korea
• /
• v.22 no.8
• /
• pp.703-710
• /
• 2003

Because of their human-like synthesized speech quality, recently Corpus-Based Text-To-Speech(CB-TTS) have been actively studied worldwide. However, due to their large size speech database (DB), their application is very restricted. In this paper we propose and evaluate three DB reduction algorithms to which are designed to solve the above drawback. The first method is based on a K-means clustering approach, which selects k-representatives among multiple instances. The second method is keeping only those unit instances that are selected during synthesis, using a domain-restricted text as input to the synthesizer. The third method is a kind of hybrid approach of the above two methods and is using a large text as input in the system. After synthesizing the given sentences, the used unit instances and their occurrence information is extracted. As next step a modified K-means clustering is applied, which takes into account also the occurrence information of the selected unit instances, Finally we compare three pruning methods by evaluating the synthesized speech quality for the similar DB reduction rate, Based on perceptual listening tests, we concluded that the last method shows the best performance among three algorithms. More than this, the results show that the last method is able to reduce DB size without speech quality looses.

• The Journal of the Acoustical Society of Korea
• /
• v.20 no.3
• /
• pp.24-33
• /
• 2001

this paper describes a baseline for an implementation of a corpus-based Korean TTS system. The conventional TTS systems using small-sized speech still generate machine-like synthetic speech. To overcome this problem we introduce the corpus-based TTS system which enables to generate natural synthetic speech without prosodic modifications. The corpus should be composed of a natural prosody of source speech and multiple instances of synthesis units. To make a phone level synthesis unit, we train a speech recognizer with the target speech, and then perform an automatic phoneme segmentation. We also detect the fine pitch period using Laryngo graph signals, which is used for prosodic feature extraction. For break strength allocation, 4 levels of break indices are decided as pause length and also attached to phones to reflect prosodic variations in phrase boundaries. To predict the break strength on texts, we utilize the statistical information of POS (Part-of-Speech) sequences. The best triphone sequences are selected by Viterbi search considering the minimization of accumulative Euclidean distance of concatenating distortion. To get high quality synthesis speech applicable to commercial purpose, we introduce a domain specific database. By adding domain specific database to general domain database, we can greatly improve the quality of synthetic speech on specific domain. From the subjective evaluation, the new Korean corpus-based TTS system shows better naturalness than the conventional demisyllable-based one.

• The Transactions of the Korea Information Processing Society
• /
• v.4 no.4
• /
• pp.923-932
• /
• 1997

In this paper, we propose a two-level hybrid signature file method(THM) to dffciently deal with large txt databases that use a term discrimination concept.In addition, we apply Yoo's clustering scheme to the two-level hybeid signature file method. The clustering schme groups similar signatures together according to the similarity of the highly discriminatiory tems so that we may achive better performance on retrival. The space-time ana-lyhtical model of the proposed two-level hybrid method is provided. Based on the analytical model and experiments, we compare it with the exsting methods, i.e. the bit-sliced method(BM), the-level method(TM), and the hybrid method(HM). As a result, we show that THM achives the best retrival performance in a large database with 100,000 records when the mumber fo matching records is less than 160.

• The Journal of the Acoustical Society of Korea
• /
• v.26 no.8
• /
• pp.403-408
• /
• 2007

In a large corpus-based speech synthesizer, a break, which is a parameter influencing the naturalness and intelligibility, is used as an important feature during a unit selection process. Japanese is a language having intonations, which ate indicated by the relative differences in pitch heights and the APs(Accentual Phrases) are placed according to the changes of the accents while a break occurs on a boundary of the APs. Although a break can be predicted by using J-ToBI(Japanese-Tones and Break Indices), which is a rule-based or statistical approach, it is very difficult to predict a break exactly due to the flexibility. Therefore, in this paper, a method is to conduct a unit search by dividing breaks into two types, such as a fixed break and a flexible break, in order to use the advantages of a large-scale corpus, which includes various types of prosodies. As a result of an experiment, the proposed unit selection method contributed itself to enhance the naturalness of synthesized speeches.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2018.05a
• /
• pp.166-168
• /
• 2018

새롭게 변형되는 대규모 악성코드들을 신속하게 탐지하기 위하여 인공지능 딥러닝을 이용한 악성코드 탐지 기법을 제안한다. 대용량의 고차원 악성코드를 저차원의 이미지로 변환하고, 딥러닝 합성곱신경망(Convolution Neural Network)을 통해 이미지의 악성코드 패턴을 학습하고 분류하였다. 본 논문에서는 악성코드 분류 모델의 성능을 검증하기 위하여 악성코드 종류별 분류 실험과 악성코드와 정상코드 분류 실험을 실시하였고 각각 97.6%, 87%의 정확도로 악성코드를 구별해 내었다. 본 논문에서 제안한 악성코드 탐지 모델은 차원 축소를 통해 10,868개(200GB)의 대규모 데이터에 대하여 10분 이내의 학습시간이 소요되어 새로운 악성코드 학습 및 대용량 악성코드 탐지를 신속하게 처리 가능함을 보였다.

• Proceedings of the KIEE Conference
• /
• 1986.07a
• /
• pp.81-85
• /
• 1986

• Journal of the Korean Chemical Society
• /
• v.59 no.2
• /
• pp.148-155
• /
• 2015

ZnO nanoplates were prepared by seed-mediated soft-solution process. Photocatalytic property of ZnO nanoplates was superior to that of conventional ZnO nanoparticles owing to the enhanced (0001) plane with large defect sites. In addition, we found that silica coating method could provide to reduce cytotoxicity of ZnO nanoplates. Finally, we have successfully synthesized for the first time large-scale synthesis of plate-type ZnO as few hundreds gram scale for industrial applications through controlling various reagents of growth solution.

• Proceedings of the Acoustical Society of Korea Conference
• /
• 1995.06a
• /
• pp.283-284
• /
• 1995

삼성종합기술원의 음성 인식, 합성 분야의 음성 데이터 베이스 구축 현황과 향후 연구 방향에 대하여 기술한다. 삼성종합기술원에서는 1989년 한국어 문음 변환기술 개발을 시작하여 그 동안 남성음, 여성음 합성 시스템을 발표하였고, 최근에는 시각장애자용 컴퓨터를 개발하여 전국 13개 시각 장애자 학교에 기정한 바 있다. 음성 인식 분야는 100 단어 내외으 소용량 화자 종속 시스템을 개발하여 키폰용 음성인식 다이얼 장치를 실용화하였다. 약 5년여에 걸친 연구 결과 자체적으로 구축하게 된 음성 DB는 크게 남, 여 합성용 DB와 인식용 DB로 요약할 수 있다. 이러한 경험을 바탕으로 향후 국내외 대학, 연구소 등과 공동연구를 통해 상품화 수준의 문음 변환기술과 대용량, 화자독립 음성인식 시스템을 개발하고자 한다. 궁극적으로는 휴대용 통역기의 요소 기술을 확보하여 제한된 영역에서 자동 통역기를 상품화하는데 이바지할 계획이다.

• The Journal of the Acoustical Society of Korea
• /
• v.17 no.4
• /
• pp.38-44
• /
• 1998

대용량의 음성합성용 데이터베이스를 용이하게 구축하기 위해 음성인식 시스템을 이용한 음소 경계 분할이 이루어지고 있다. 그러나 자동 분할 결과를 직접 이용하여 합성음 을 생성할 경우 음소 경계 에러로 인하여 접합 왜곡이 많이 발생하게 된다. 이러한 문제를 해결하기 위해서, 본 연구에서는 단위 접합시 경계 에러를 고려하여 적합한 접합 위치를 찾 고자 하였다. 여기서 적합한 접합 위치는 스펙트럼의 불연속이 최소화된 접합점을 의미한다. 합성음에 대한 MOS(Mean Opinion Score) 테스트와 스펙트로그램(spectrogram)의 모양을 비교하므로써 제안된 방법의 성능을 평가하였다. 제안된 방법은 두 단계로 이루어져 있다. 첫째, 레퍼런스 패턴(reference pattern)과 두 개의 테스트 패턴(test pattern)을 선택하는 단 계와, 둘째, 앞과 뒤 테스트 패턴 사이의 적합한 접합위치를 찾는 단계이다. 본 연구에서는 패턴 사이의 스펙트로그램 비교를 위해 켑스트럼(cepstrum) 피라미터와 패턴 분류기 (pattern classifier)인 DTW(Dynamic Time Warping) 알고리즘을 사용하였다. 제안된 알고 리즘을 평가한 청취 테스트의 결과에서 제안된 알고리즘을 적용하여 합성된 합성음의 음질 이 자동 분절로 생성된 단위를 그대로 이용한 경우의 음질보다 우수함을 보였다.

• Journal of the Korea Society of Computer and Information
• /
• v.15 no.7
• /
• pp.91-98
• /
• 2010

Large corpus-based concatenating Text-to-Speech (TTS) systems can generate natural synthetic speech without additional signal processing. Because the improvements in the natualness, personality, speaking style, emotions of synthetic speech need the increase of the size of speech DB, it is necessary to prune the redundant speech segments in a large speech segment DB. In this paper, we propose a new method to construct a segmental speech DB for the Korean TTS system based on a clustering algorithm to downsize the segmental speech DB. For the performance test, the synthetic speech was generated using the Korean TTS system which consists of the language processing module, prosody processing module, segment selection module, speech concatenation module, and segmental speech DB. And MOS test was executed with the a set of synthetic speech generated with 4 different segmental speech DBs. We constructed 4 different segmental speech DB by combining CM1(or CM2) tree clustering method and full DB (or reduced DB). Experimental results show that the proposed method can reduce the size of speech DB by 23% and get high MOS in the perception test. Therefore the proposed method can be applied to make a small sized TTS.