• Phonetics and Speech Sciences
• /
• v.13 no.3
• /
• pp.71-78
• /
• 2021

Most neural-network-based speech synthesis models utilize neural vocoders to convert mel-scaled spectrograms into high-quality, human-like voices. However, neural vocoders combined with mel-scaled spectrogram prediction models demand considerable computer memory and time during the training phase and are subject to slow inference speeds in an environment where GPU is not used. This problem does not arise in linear spectrogram prediction models, as they do not use neural vocoders, but these models suffer from low voice quality. As a solution, this paper proposes a Tacotron 2 and Transformer-based linear spectrogram prediction model that produces high-quality speech and does not use neural vocoders. Experiments suggest that this model can serve as the foundation of a high-quality text-to-speech model with fast inference speed.

• Proceedings of the IEEK Conference
• /
• 1999.06a
• /
• pp.699-707
• /
• 1999

It cannot be argued that speech is the most natural interfacing tool between men and machines. In order to realize acceptable speech interfaces, highly advanced speech recognizers and synthesizers are inevitable. Text-to-Speech(TTS) technology has been attracting a lot of interest among speech engineers because of its own benefits. Namely, the possible application areas of talking computers, emergency alarming systems in speech, speech output devices fur speech-impaired, and so on. Hence, many researchers have made significant progresses in the speech synthesis techniques in the sense of their own languages and as a result, the quality of currently available speech synthesizers are believed to be acceptable to normal users. These are partly why the MPEG group had decided to include the TTS technology as one of its MPEG-4 functionalities. ETRI has made major contributions to the current MPEG-4 TTS among various MPEG-4 functionalities. They are; 1) use of original prosody for synthesized speech output, 2) trick mode functions fer general users without breaking synthesized speech prosody, 3) interoperability with Facial Animation(FA) tools, and 4) dubbing a moving/animated picture with lib-shape pattern information.

• Speech Sciences
• /
• v.2
• /
• pp.25-44
• /
• 1997

When listening the various speech synthesis systems developed and being used in our country, we find that though the quality of these systems has improved, they lack naturalness. Moreover, since the voice color of these systems are limited to only one recorded speech DB, it is necessary to record another speech DB to create different voice colors. 'Voice Color' is an abstract concept that characterizes voice personality. So speech synthesis systems need a voice color control function to create various voices. The aim of this study is to examine several factors of voice color control rules for the text-to-speech system which makes natural and various voice types for the sounding of synthetic speech. In order to find such rules from natural speech, glottal source parameters and frequency characteristics of the vocal tract for several voice colors have been studied. In this paper voice colors were catalogued as: deep, sonorous, thick, soft, harsh, high tone, shrill, and weak. For the voice source model, the LF-model was used and for the frequency characteristics of vocal tract, the formant frequencies, bandwidths, and amplitudes were used. These acoustic parameters were tested through multiple regression analysis to achieve the general relation between these parameters and voice colors.

• MALSORI
• /
• no.43
• /
• pp.73-88
• /
• 2002

In this paper, for natural facial synthesis, lip-synch algorithm based on key-frame method using RBF(radial bases function) is presented. For lips synthesizing, we make viseme range parameters from phoneme and its duration information that come out from the text-to-speech(TTS) system. And we extract viseme information from Av DB that coincides in each phoneme. We apply dominance function to reflect coarticulation phenomenon, and apply bilinear interpolation to reduce calculation time. At the next time lip-synch is performed by playing the synthesized images obtained by interpolation between each phonemes and the speech sound of TTS.

• 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.

• MALSORI
• /
• no.53
• /
• pp.75-92
• /
• 2005

Prosody Generation Based on C-ToBI Representation for Text-to-SpeechSeungwon Kim, Yu Zheng, Gary Geunbae Lee, Byeongchang KimProsody modeling is critical in developing text-to-speech (TTS) systems where speech synthesis is used to automatically generate natural speech. In this paper, we present a prosody generation architecture based on Chinese Tone and Break Index (C-ToBI) representation. ToBI is a multi-tier representation system based on linguistic knowledge to transcribe events in an utterance. The TTS system which adopts ToBI as an intermediate representation is known to exhibit higher flexibility, modularity and domain/task portability compared with the direct prosody generation TTS systems. However, the cost of corpus preparation is very expensive for practical-level performance because the ToBI labeled corpus has been manually constructed by many prosody experts and normally requires a large amount of data for accurate statistical prosody modeling. This paper proposes a new method which transcribes the C-ToBI labels automatically in Chinese speech. We model Chinese prosody generation as a classification problem and apply conditional Maximum Entropy (ME) classification to this problem. We empirically verify the usefulness of various natural language and phonology features to make well-integrated features for ME framework.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.23 no.6
• /
• pp.1443-1453
• /
• 1998

This paper presents the speech synthesis engine that converts the character strings kept in a computer memory into the synthesized speech sounds with enhancing the intelligibility and the naturalness by adapting the waveform processing method. The speech engine using demisyllable speech segments receives command streams for pitch modification, duration and energy control. The command based engine isolates the high level processing of text normalization, letter-to-sound and the lexical analysis and the low level processing of signal filtering and pitch processing. The TTS(Text-to-Speech) system implemented by using the speech synthesis engine has three independent object modules of the Text-Normalizer, the Commander and the said Speech Synthesis Engine those of which are easily replaced by other compatible modules. The architecture separating the high level and the low level processing has the advantage of the expandibility and the portability because of the mix-and-match nature.

• Journal of the Korean Institute of Telematics and Electronics S
• /
• v.35S no.7
• /
• pp.95-102
• /
• 1998

This paper proposes a synchronization method of synthetic facial iamge sequences and synthetic speech. The LP-PSOLA synthesizes the speech for each demi-syllable. We provide the 3,040 demi-syllables for unlimited synthesis of the Korean speech. For synthesis of the Facial image sequences, the paper defines the total 11 fundermental patterns for the lip shapes of the Korean consonants and vowels. The fundermental lip shapes allow us to pronounce all Korean sentences. Image synthesis method assigns the fundermental lip shapes to the key frames according to the initial, the middle and the final sound of each syllable in korean input text. The method interpolates the naturally changing lip shapes in inbetween frames. The number of the inbetween frames is estimated from the duration time of each syllable of the synthetic speech. The estimation accomplishes synchronization of the facial image sequences and speech. In speech synthesis, disk memory is required to store 3,040 demi-syllable. In synthesis of the facial image sequences, however, the disk memory is required to store only one image, because all frames are synthesized from the neutral face. Above method realizes synchronization of system which can real the Korean sentences with the synthetic speech and the synthetic facial iage sequences.

• The Journal of the Acoustical Society of Korea
• /
• v.14 no.3
• /
• pp.49-60
• /
• 1995

This paper is a study on the design and implementation of the Korean Text-to-Speech system which is used for a small and simple system. In this paper, a parameter synthesis method is chosen for speech syntheiss method, we use PARCOR(PARtial autoCORrelation) coefficient which is one of the LPC analysis. And we use phoneme for synthesis unit which is the basic unit for speech synthesis. We use PARCOR, pitch, amplitude as synthesis parameter of voice, we use residual signal, PARCOR coefficients as synthesis parameter of unvoice. In this paper, we could obtain the 60% intelligibility by using the residual signal as excitation signal of unvoiced sound. The result of synthesis experiment, synthesis of a word unit is available. The controlling of phoneme duration is necessary for synthesizing of a sentence unit. For setting up the synthesis system, PC 486, a 70[Hz]-4.5[KHz] band pass filter for speech input/output, amplifier, and TMS320C30 DSP board was used.

• MALSORI
• /
• no.56
• /
• pp.225-235
• /
• 2005

Speech unit concatenation with a large database is presently the most popular method for speech synthesis. In this approach, the mismatches at the unit boundaries are unavoidable and become one of the reasons for quality degradation. This paper proposes an algorithm to reduce undesired discontinuities between the subsequent units. Optimal matching points are calculated in two steps. Firstly, the fullback-Leibler distance measurement is utilized for the spectral matching, then the unit sliding and the overlap windowing are used for the waveform matching. The proposed algorithm is implemented for the corpus-based unit concatenating Korean text-to-speech system that has an automatically labeled database. Experimental results show that our algorithm is fairly better than the raw concatenation or the overlap smoothing method.