• MALSORI
• /
• no.50
• /
• pp.41-50
• /
• 2004

Recently, there has been considerable progress in the natural language processing and digital signal processing components and this progress has led to the improved synthetic speech qualify of many commercial TTS systems. But there still remain many obstacles to overcome for the practical application of TTS. To resolve the problems, the cooperative research among the related areas is highly required and a common Korean TTS platform is essential to promote these activities. This platform offers a general framework for building Korean speech synthesis systems and a full C/C++ source for modules supports to implement and test his own algorithm. In this paper we described the aspect of a Korean TTS platform to be developed and a developing plan.

• Phonetics and Speech Sciences
• /
• v.10 no.1
• /
• pp.39-48
• /
• 2018

A typical statistical parametric speech synthesis (text-to-speech, TTS) system consists of separate modules, such as a text analysis module, an acoustic modeling module, and a speech synthesis module. This causes two problems: 1) expert knowledge of each module is required, and 2) errors generated in each module accumulate passing through each module. An end-to-end TTS system could avoid such problems by synthesizing voice signals directly from an input string. In this study, we implemented an end-to-end Korean TTS system using Google's Tacotron, which is an end-to-end TTS system based on a sequence-to-sequence model with attention mechanism. We used 4392 utterances spoken by a Korean female speaker, an amount that corresponds to 37% of the dataset Google used for training Tacotron. Our system obtained mean opinion score (MOS) 2.98 and degradation mean opinion score (DMOS) 3.25. We will discuss the factors which affected training of the system. Experiments demonstrate that the post-processing network needs to be designed considering output language and input characters and that according to the amount of training data, the maximum value of n for n-grams modeled by the encoder should be small enough.

• 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.

• The Journal of the Convergence on Culture Technology
• /
• v.6 no.2
• /
• pp.509-514
• /
• 2020

The conventional TTS system consists of several modules, including text preprocessing, parsing analysis, grapheme-to-phoneme conversion, boundary analysis, prosody control, acoustic feature generation by acoustic model, and synthesized speech generation. But TTS system with deep learning is composed of Text2Mel process that generates spectrogram from text, and vocoder that synthesizes speech signals from spectrogram. In this paper, for the optimal Korean TTS system construction we apply Tacotron2 to Tex2Mel process, and as a vocoder we introduce the methods such as WaveNet, WaveRNN, and WaveGlow, and implement them to verify and compare their performance. Experimental results show that WaveNet has the highest MOS and the trained model is hundreds of megabytes in size, but the synthesis time is about 50 times the real time. WaveRNN shows MOS performance similar to that of WaveNet and the model size is several tens of megabytes, but this method also cannot be processed in real time. WaveGlow can handle real-time processing, but the model is several GB in size and MOS is the worst of the three vocoders. From the results of this study, the reference criteria for selecting the appropriate method according to the hardware environment in the field of applying the TTS system are presented in this paper.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2013.11a
• /
• pp.1238-1240
• /
• 2013

본 논문에서는 AAC기술과 TTS기술을 조사 및 분석하여 스마트폰의 가장기본적인 기능인 전화와 문자전달 기능을 일반적으로 의사소통이 어려운 언어장애인들에게 보완 대체 의사소통의 수단을 제공하는 MLS시스템을 제안하고, Text to Speech기능과 의사소통기능, TTS전화기능, 설정기능을 설계하였다.

• MALSORI
• /
• no.46
• /
• pp.51-64
• /
• 2003

In order to produce high quality synthesized speech, it is very important to get an accurate grapheme-to-phoneme conversion and prosody model from texts using natural language processing. Robust preprocessing for non-Korean characters should also be required. In this paper, we analyzed Korean texts using a morphological analyzer, part-of-speech tagger and syntactic chunker. We present a new grapheme-to-phoneme conversion method for Korean using a hybrid method with a phonetic pattern dictionary and CCV (consonant vowel) LTS (letter to sound) rules, for unlimited vocabulary Korean TTS. We constructed a prosody model using a probabilistic method and decision tree-based method. The probabilistic method atone usually suffers from performance degradation due to inherent data sparseness problems. So we adopted tree-based error correction to overcome these training data limitations.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2013.11a
• /
• pp.1490-1492
• /
• 2013

최근 들어 스마트폰과 같은 휴대용 장치에 안드로이드 운영체제가 널리 탑재되면서, 안드로이드용 애플리케이션이 활발히 개발되고 있다. 본 논문에서는 텍스트를 음성으로 변환하는 안드로이드용 TTS(Text-to-Speech) 애플리케이션을 개발한다. 기존의 유사 TTS 애플리케이션들과 달리, 본 논문에서 개발한 애플리케이션은 사용자가 직접 입력한 임의의 텍스트를 음성으로 읽어주는 기능과 미리 저장되어 있는 전자책을 불러와 음성으로 읽어주는 기능을 모두 제공한다. 또한 한글과 영어를 모두 지원하기 때문에, 한글 텍스트와 영어 텍스트를 모두 사용할 수 있다. 따라서 본 애플리케이션은 여러 사용자에 의해 다양한 용도로 사용될 수 있다.

• The Journal of the Convergence on Culture Technology
• /
• v.7 no.1
• /
• pp.640-645
• /
• 2021

The deep learning based end-to-end TTS system consists of Text2Mel module that generates spectrogram from text, and vocoder module that synthesizes speech signals from spectrogram. Recently, by applying deep learning technology to the TTS system the intelligibility and naturalness of the synthesized speech is as improved as human vocalization. However, it has the disadvantage that the inference speed for synthesizing speech is very slow compared to the conventional method. The inference speed can be improved by applying the non-autoregressive method which can generate speech samples in parallel independent of previously generated samples. In this paper, we introduce FastSpeech, FastSpeech 2, and FastPitch as Text2Mel technology, and Parallel WaveGAN, Multi-band MelGAN, and WaveGlow as vocoder technology applying non-autoregressive method. And we implement them to verify whether it can be processed in real time. Experimental results show that by the obtained RTF all the presented methods are sufficiently capable of real-time processing. And it can be seen that the size of the learned model is about tens to hundreds of megabytes except WaveGlow, and it can be applied to the embedded environment where the memory is limited.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2014.11a
• /
• pp.806-808
• /
• 2014

현재까지의 음성 출력 시스템은 텍스트를 있는 그대로 읽어 주는 것에 불과했다. 단위, 약어의 경우 알파벳을 그대로 읽어 주게 되어 그 본래의 의미를 제대로 파악하기 어려웠다. 본 연구에서는 단위나 약어의 본래의 의미를 찾아서 풀어서 음성 변환해 주는 방법을 제안함으로써 시각 장애인에게도 텍스트의 정확한 정보를 전달할 수 있다는 장점이 있다.

• 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.