• Proceedings of the Korean Society for Language and Information Conference
• /
• 2007.11a
• /
• pp.405-414
• /
• 2007

This paper describes a preliminary work on prosody modeling aspect of a text-to-speech system for Thai. Specifically, the model is designed to predict symbolic markers from text (i.e., prosodic phrase boundaries, accent, and intonation boundaries), and then using these markers to generate pitch, intensity, and durational patterns for the synthesis module of the system. In this paper, a novel method for annotating the prosodic structure of Thai sentences based on dependency representation of syntax is presented. The goal of the annotation process is to predict from text the rhythm of the input sentence when spoken according to its intended meaning. The encoding of the prosodic structure is established by minimizing speech disrhythmy while maintaining the congruency with syntax. That is, each word in the sentence is assigned a prosodic feature called strength dynamic which is based on the dependency representation of syntax. The strength dynamics assigned are then used to obtain rhythmic groupings in terms of a phonological unit called foot. Finally, the foot structure is used to predict the durational pattern of the input sentence. The aforementioned process has been tested on a set of ambiguous sentences, which represents various structural ambiguities involving five types of compounds in Thai.

• Speech Sciences
• /
• v.9 no.3
• /
• pp.25-34
• /
• 2002

One of the most important factor in the performance of Korean text-to-speech system is the prediction of accentual and intonational phrase boundary. The previous method of prediction shows only the 75-85% which is not proper in the practical and commercial system. Therefore, more accurate prediction must be needed in the practical system. In this study, we propose the simple and more accurate method of the prediction of AP, IP.

• Journal of the Korean Institute of Telematics and Electronics B
• /
• v.28B no.10
• /
• pp.755-760
• /
• 1991

In this paper, we propose a new formant locus overlapping method which can effectively enhance a naturalness of synthetic speech produced by ddemisyllable based Korean text-to-speech system. At first, Korean demisyllables are divided into several number of segments which have linear formant transition characteristics. Then, database, which is composed of start point and length of each formant segments, is provided. When we synthesize speech with these demisyllable database, we concatenate each formant locus by using a proposed overlapping method which can closely simulate haman articulation mechanism. We have implemented a Korean text-to-speech system by using this method and proved that the formant loci of synthetic speech are similar to those of the natural speech. Finally, we could illustrate that the resulting spectrograms of proposed method are more similar to natural speech than those of conventional method.

• 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.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.14 no.8
• /
• pp.3473-3487
• /
• 2020

In this paper, we present a new approach to creating speech animation with emotional expressions using a small set of example models. To generate realistic facial animation, two example models called key visemes and expressions are used for lip-synchronization and facial expressions, respectively. The key visemes represent lip shapes of phonemes such as vowels and consonants while the key expressions represent basic emotions of a face. Our approach utilizes a text-to-speech (TTS) system to create a phonetic transcript for the speech animation. Based on a phonetic transcript, a sequence of speech animation is synthesized by interpolating the corresponding sequence of key visemes. Using an input parameter vector, the key expressions are blended by a method of scattered data interpolation. During the synthesizing process, an importance-based scheme is introduced to combine both lip-synchronization and facial expressions into one animation sequence in real time (over 120Hz). The proposed approach can be applied to diverse types of digital content and applications that use facial animation with high accuracy (over 90%) in speech recognition.

• English Language & Literature Teaching
• /
• v.12 no.2
• /
• pp.87-105
• /
• 2006

Style is considered a pivotal construct in sociolinguistic variation studies. While previous studies have examined style in traditional forms of language such as speech, very little research has examined new and emerging styles such as computer-mediated discourse. Thus, the present study attempts to investigate style in the online communication mode of chat. In so doing, the study compares text-based online chat with speech and writing. Online chat has been previously described as a hybrid form of language that is close to speech. Here, the exact nature of online chat is elucidated by focusing on contraction use. Differential acquisition of stylistic variation is also examined according to English learning background. The empirical component consists of data from Korean speakers of English. Data is taken from a written summary, an oral interview, and a text-based online chat session. A multivariate analysis was conducted. Results indicate that online chat is indeed a hybrid form that is difficult to delineate from speech and writing. Text-based online chat shows a somewhat similar rate of contraction to speech, which confirms its hybridity.. Lastly, some implications of the study are given in terms of the learning and acquisition of style in general and in online contextual modes.

• Phonetics and Speech Sciences
• /
• v.13 no.4
• /
• pp.47-53
• /
• 2021

Autoregressive Text-to-Speech (TTS) models suffer from inference instability and slow inference speed. Inference instability occurs when a poorly predicted sample at time step t affects all the subsequent predictions. Slow inference speed arises from a model structure that forces the predicted samples from time steps 1 to t-1 to predict the sample at time step t. In this study, an end-to-end non-autoregressive fast text-to-speech model is suggested as a solution to these problems. The results of this study show that this model's Mean Opinion Score (MOS) is close to that of Tacotron 2 - WaveNet, while this model's inference speed and stability are higher than those of Tacotron 2 - WaveNet. Further, this study aims to offer insight into the improvement of non-autoregressive models.

• MALSORI
• /
• no.53
• /
• pp.61-74
• /
• 2005

In the present study, the intelligibility of the synthesized speech sounds was evaluated by using the psycholinguistic and fMRI techniques. In order to see the difference in recognizing words between the natural and synthesized speech sounds, word regularity and word frequency were varied. The results of Experiment1 and Experiment2 showed that the intelligibility difference of the synthesized speech comes from word regularity. In the case of the synthesized speech, the regular words were recognized slower than the irregular words, and there was smaller activation of the auditory areas in brain for the regular words than for the irregular words.

• Phonetics and Speech Sciences
• /
• v.12 no.2
• /
• pp.29-37
• /
• 2020

Recently, the neural network-based deep learning algorithm has dramatically improved performance compared to the classical Gaussian mixture model based hidden Markov model (GMM-HMM) automatic speech recognition (ASR) system. In addition, researches on end-to-end (E2E) speech recognition systems integrating language modeling and decoding processes have been actively conducted to better utilize the advantages of deep learning techniques. In general, E2E ASR systems consist of multiple layers of encoder-decoder structure with attention. Therefore, E2E ASR systems require data with a large amount of speech-text paired data in order to achieve good performance. Obtaining speech-text paired data requires a lot of human labor and time, and is a high barrier to building E2E ASR system. Therefore, there are previous studies that improve the performance of E2E ASR system using relatively small amount of speech-text paired data, but most studies have been conducted by using only speech-only data or text-only data. In this study, we proposed a semi-supervised training method that enables E2E ASR system to perform well in corpus in different domains by using both speech or text only data. The proposed method works effectively by adapting to different domains, showing good performance in the target domain and not degrading much in the source domain.

• Proceedings of the Acoustical Society of Korea Conference
• /
• 1993.06a
• /
• pp.98-101
• /
• 1993

본 논문은 한국전자통신연구소 음성응용연구실에서 개발한 문자-음성변환 시스팀(Text-to-Speech Conversion System)의 음질개선 연구의 일환으로 Phoneme-Balanced Words 110개에 대해서 개선전 시스팀(V.1)과 개선 후 시스팀(v.2)을 대상으로 각각 실시한 명료도 실험결과에 대하여 기술하고 있다. 본 실험의 목적은 연구개발자 입장에서 합성음 개선에 대한 정량적 성과 및 문제점 파악을 위한 진단형 평가이며 남자 5명, 여자 5명을 대상으로 1회 실시한 청취 실험결과 V.1에 대해서는 최저 37.3%(41개) ~ 최고 55.5%(61개)이고, V.2에 대해서는 최고 39.1%(43개) ~ 최고 60.9%(67개) 결과를 얻었다.