• Phonetics and Speech Sciences
• /
• v.16 no.1
• /
• pp.67-76
• /
• 2024

Recent advances in text-to-speech (TTS) technology have significantly improved the quality of synthesized speech, reaching a level where it can closely imitate natural human speech. Especially, TTS models offering various voice characteristics and personalized speech, are widely utilized in fields such as artificial intelligence (AI) tutors, advertising, and video dubbing. Accordingly, in this paper, we propose a one-shot multi-speaker TTS system that can ensure acoustic diversity and synthesize personalized voice by generating speech using unseen target speakers' utterances. The proposed model integrates a speaker encoder into a TTS model consisting of the FastSpeech2 acoustic model and the HiFi-GAN vocoder. The speaker encoder, based on the pre-trained RawNet3, extracts speaker-specific voice features. Furthermore, the proposed approach not only includes an English one-shot multi-speaker TTS but also introduces a Korean one-shot multi-speaker TTS. We evaluate naturalness and speaker similarity of the generated speech using objective and subjective metrics. In the subjective evaluation, the proposed Korean one-shot multi-speaker TTS obtained naturalness mean opinion score (NMOS) of 3.36 and similarity MOS (SMOS) of 3.16. The objective evaluation of the proposed English and Korean one-shot multi-speaker TTS showed a prediction MOS (P-MOS) of 2.54 and 3.74, respectively. These results indicate that the performance of our proposed model is improved over the baseline models in terms of both naturalness and speaker similarity.

• Journal of the Korean Society of Laryngology, Phoniatrics and Logopedics
• /
• v.15 no.2
• /
• pp.122-127
• /
• 2004

Background and Objectives : Many studies have described and analyzed singer's formant and it has been shown that the epilaryngeal tube in the human airway is responsible for vocal ring, or the singer's formant. A similar phenomenon produced by trained singers in their speech led some authors to examine the speaker's ring. This study was designed to analyze the speaking voice of the singers and speaker's ring. Baterials and Methods : Ten tenors, fifteen baritones, fifteen sopranos and ten mezzo sopranos attending the music college, department of vocal music were chosen for this study. Fifteen male and fifteen female untrained normal speakers were chosen for control group. Each subject was asked to produce a sample of a sustained spoken vowel /ah/ sound for at least five seconds and read sentence 'Kaeul'. The sound data was analyzed using the Fast Fourier Transform(FFT) - based power spectrum, Long term average(LTA) power spectrum using the FFT algorithm of the Computerized Speech Lab(CSL, Kay elemetrics, Model 4300B, USA). Statistical analysis was performed using the Mann-Whitney test of the Statistical Package for Social Sciences(SPSS). Results : For LTA Power spectrum of/ah/ sound, a significant increase was seen in the 2,500-3,500Hz region(p<0.01) in four trained singer group compared with untrained speaker group, and a significant increase in the 9,000-10,000Hz region(p<0.01) in soparano group. Similarly, in sentence 'Kaeul', there was a significant increase in energy in the tenor, baritone, mezzo soprano group compared with the untrained speaker group in the 2,500-3,500Hz region(p<0.01), and a significant increase in all frequency region(p<0.01) in the soprano group. Conclusions : The LTA power spectrum suggests that trained singers group show more energy concentration in the 'singer's formant' region in the speaking voice, and authors believe this region to be the 'speaker's ring'. Further research is needed on the effect of singing training on the resonance of the speaking voice.

• Korean Journal of Cognitive Science
• /
• v.17 no.4
• /
• pp.303-321
• /
• 2006

According to Harris'(1966) distributional hypothesis, understanding the meaning of a word is thought to be dependent on its context. Under this hypothesis about human language ability, this paper proposes a computational model for native speaker's language processing mechanism concerning word sense disambiguation, based on two sets of experiments. Among the three computational models discussed in this paper, namely, the logic model, the probabilistic model, and the probabilistic inference model, the experiment shows that the logic model is first applied fer semantic disambiguation of the key word. Nexr, if the logic model fails to apply, then the probabilistic model becomes most relevant. The three models were also compared with the test results in terms of Pearson correlation coefficient value. It turns out that the logic model best explains the human decision behaviour on the ambiguous words, and the probabilistic inference model tomes next. The experiment consists of two pans; one involves 30 sentences extracted from 1 million graphic-word corpus, and the result shows the agreement rate anong native speakers is at 98% in terms of word sense disambiguation. The other pm of the experiment, which was designed to exclude the logic model effect, is composed of 50 cleft sentences.

• Journal of the Korean BIBLIA Society for library and Information Science
• /
• v.13 no.2
• /
• pp.221-242
• /
• 2002

This study is about special programs for elementary school library media centers in Texas in the United States, that can be used as a way of activating the school library media center programs. Various special programs of the seven elementary school library media centers in Texas were selected for case studies and the types, purposes. frameworks, effects of the programs and the subjects were analyzed. Special programs were provided for exhibits and displays, featured speakers or entertainers, learning centers, demonstrations, and media in conjunction with book fair and curriculum. And the purposes of the programs were to support and enhance the curriculum. to provide personal enrichment and to provide professional development for teachers. The frameworks for the program were the special target group, curriculum needs, special theme. interest and entertainment within available time periods. The benefits of the programs were the highlights of the media center and the media specialist, increasing of the school library media center usage. classroom support and enrichment, broadening students' and teachers' interests. encouragement of the development of lifelong learners and the establishment of the good public relations. Therefore, we have to consider the development of the special programs as a method of activating the school library in Korea and it should be brought with the concern and support from the principals, teachers, parents, and community members.

• Journal of the Korean Society for Marine Environment & Energy
• /
• v.5 no.3
• /
• pp.28-34
• /
• 2002

As the technology of Speech Recognition(SR) and Text-To-Speech(TTS) develops rapidly, voice control and guidance system is thought to be very helpful for safe navigation. But Voice Control and Guidance System(VCGS) is not yet so popularly included in Navigation Supporting System(NSS). The main reason of this is that VCGS is so complicated and user-unfriendly that navigation officers hesitate to use VCGS. Frequent errors in operating VCGS due to low rate of SR are another reason. To make VCGS more practicable for safe navigation, we design the user-friendly VCGS. Firstly, by using interviews we survey functions and procedures that navigation officers want to be included in VCGS. Secondly, to raise the rate of SR, we tun the environmental noise in bridge and to reduce the errors due to low rate of SR in operating VCGS, we design the functions of self-correction. Also we apply a user-independent SR engine so that procedures of teaming of speakers is basically not necessary. Using simulator experiments the functions and procedures of the user-friendly YCGS for safe navigation are evaluated and the results of evaluation are fed back to the design. As a result, we can design the VCGS more helpful for safe navigation. In this paper, we describe the features of the user-friendly VCGS for safe navigation and discuss the results of simulator experiments.

• Korean Journal of Community Nutrition
• /
• v.13 no.6
• /
• pp.818-828
• /
• 2008

This study was conducted to investigate actual conditions and needs on breastfeeding education for pregnant women in health centers for the development of a breastfeeding educational program. The questionnaires were mailed to 245 health centers and 148 (60.4%) questionnaires were returned. 91% of the health centers had breastfeeding education for pregnant women. 66% of them operated breastfeeding education as one part of other health programs. About 76% of supervisors and 64% of educators were the nurses. The teaching methods frequently used were lectures (30.5%), giving out booklets and leaflets (22.6%), demonstration and practice (21.5%), personal counseling (13.3%), and others. The teaching materials used were materials of outside speakers (39.7%), materials of development oneself (19.0%), materials of academic association or institute (14.8%), and others. The subjects which educators taught were the benefits of breastfeeding (16.7%), breastfeeding techniques (15.8%), caring for breasts before and after delivery (15.1%), nutritional management for lactating women (14.2%), coping strategies for the difficult situation of breastfeeding (13.3%), and others. Those were different from each other according to the educators' general characteristics. Success factors of education were increased motivation for breastfeeding (52.8%), practice (22.6%), professional's lecture (11.3%) and others. The failure factors of education were the ineffectiveness of the lecture method (69.2%), lack of education for supporters (15.4%) and lack of standardized education (15.4%). The most important barrier of education was the lack of a standardized breastfeeding educational programs (43.9%). The most effective teaching methods that educators thought were demonstration and practice (24.0%). The educators thought they need the tools and space for practice (28.2%), a standardized breastfeeding educational program (26.9%), and the human resources (24.4%) for effective education. Subjects that educators thought important for education were the breastfeeding techniques, benefits of breastfeeding, caring for breasts before and after delivery, nutritional management for lactating women, coping strategies for the difficult situation of breastfeeding, rooming system after delivery, ways to assess mother's milk quantity, introducing successful cases of breastfeeding in rank order. To promote the effectiveness of breastfeeding education, standardized breastfeeding educational programs, diverse teaching materials, space and tools, and human resources are needed.

• Korean journal of communication and information
• /
• v.63
• /
• pp.26-52
• /
• 2013

This study started from the questions on how the selection of the 'New 7 Wonders of Nature' was repeatedly reported, and whose voice was emphasized on which perspective in regional daily newspapers. This is a subject that received the attention of the media with the tag of 'fraud on a national scale'. The results of this study showed that there was no difference among newspapers in all frames excluding the frame of public confidence in relation to the N7W Foundation. Furthermore, after examining the change before and after six media companies' conclusion of MOUs, a significant difference was discovered between newspapers concerning the frame of public confidence pertaining to the N7W Foundation. However, no significant difference was discovered in all frames after the MOUs were concluded. The tone of the press mostly displayed a positive or supporting perspective, and almost no negative or critical reports were published. Second, 68.0% of all the articles displayed use direct quotes. The speakers of direct quotes were mostly from official organizations that initiated the selection event for the 'New 7 Wonders of Nature', and almost no civilians or representatives from civil social groups were quoted. On the other hand, direct quotes of anonymous sources were used rather frequently, and most of these sources were confirmed to be official sources and related parties that initiated the selection event for the 'New 7 Wonders of Nature'. As a result, regional daily newspapers displayed support and advocated for the voices of the entities that initiated the selection event for the 'New 7 Wonders of Nature', including the district government, by emphasizing and repeating their perspectives in the newspaper coverage.

• The Journal of the Convergence on Culture Technology
• /
• v.6 no.1
• /
• pp.321-326
• /
• 2020

This paper is to analyze the language education policy in the context of multilingualism. As the majority of the population are multilingual, language policy should be centered on the multilingual speakers as the norm, and multilingual language policy is the best route which we can follow as a language policy in education. The motivation and legitimacy of the multilingual policies are suggested in terms of 6 different perspectives: identity, sustainability, equity, World Englishes, machine translation, and Universal Grammar (UG). As a model of language policy, the English-Plus (i.e., English+n) policy and similarly the Korean-Plus (i.e., Korean+n) policy are suggested to be the most appropriate language policies in the field of education in America and Korea respectively. These plus policies aim at bilingual fluency in both the native language and other foreign languages that are constitutive of the multilingualism of the country in which the bilingualism is treated as a variant of multilingualism. In a period of convergence and diversity in the 4^th Industrial Revolution, language diversity and multilingual policy should be considered as a right to be protected or as a resource to be conserved rather than as a problem to be solved.

• The Journal of the Acoustical Society of Korea
• /
• v.34 no.2
• /
• pp.138-145
• /
• 2015

Although it is well-known that environmental noise can lead to hearing loss in individuals, the true extent of subway noise effects in the general population remains poorly understood. The purpose of the present study is to see changes of acoustic reflex thresholds and speech perception scores when passengers listen to music presented from their personal listening device in the subway. Forty subjects with normal hearing participated being divided into two groups, experimental and control groups. As a baseline, all subjects were measured by acoustic reflex thresholds in five test frequencies and Korean speech perception in noise (KSPIN) test at 0 and -5 dB SNR. In the experiment, the control group read newspaper or magazine in the subway noise, whereas the experimental group listened to music presented from their smartphone under the subway noise through speakers at 73.45 dBA for 60 min. After completing the experiment, two groups also conducted both acoustic reflex thresholds and KSPIN tests in the same condition as the baseline. The results showed that there was a significant difference of correct percent in speech-in-noise test between experimental and control groups and of that between two signal-to-noise ratios, which means the double noise exposure of 60 min might cause some degradation of speech perception when noise increases compared to only subway noise condition that was not statistically significant difference. We concluded that a risk of some degraded speech perception ability would be expected when passengers have a habit of listening to music in the subway noisy situation for a long duration.

• Sijohaknonchong
• /
• v.19 no.1
• /
• pp.85-106
• /
• 2003

Recently, the sonnet and the shijo, each representing French and Korean formal poetries, are tend to be read with the eyes only, as were more accustomed to written literature. But even after almost three millennia of written literature and increased use of digitalized poems, poetry retains its appeal to the ear as well as to the eye. To read a poem only by eyes might be wrong because it is designed to be read aloud by mouth and understood by ear, and will decrease the aesthetic sense otherwise. It is essential to find the right way to recite a poem in this dramatically changed society, and is especially important when many shijos are changing into digitalized forms to adapt the new wave of our society. The sonnet and the shijo emphasize the importance of the harmony of sounds and rhythms with certain structure, and have their own prosodies. The emotions of the speaker in poems are expressed with words. When they are pronounced. each phoneme has its own phonemic characteristics. When comparing the The Broken Bell(Baudelaire) and Chopoong ga (Jong Seo Kim) in terms of prosody and phonetics. the speakers emotions are closely related with the phonetic structure of each word. In The Broken Bell, the phonetic value of rhymes, repeated phonemes, concentration of front and back vowels. rhythms of onesyllable words shape the overall image of this poem describing the productivity of bells as appose to the sterility of the soul. Chopoong ga also shows the determined and strong will of the speaker by frequent glottalized sounds. distribution and concentration of certain vowels. and frequent use of plosives. As you see in these examples, phones, beats, and rhythms are not the mere transmitter of meaning but possess their expressive values of their own and should be the first to be considered when reciting a poem.