Pattern Analysis of Voice Onset and Offset in Normal Adults Using High-Speed Digital Imaging: The Role of Arytenoid Cartilage Movements

최성희,오치선,최철희 한국언어청각임상학회 2015 Communication Sciences and Disorders Vol.20 No.4

Objectives: Voice onset and offset include dynamic changes of aerodynamic and biomechanical properties showing irregular and chaotic voice signals during phonation. This preliminary study investigated voice onset and offset patterns based on gender using high-speed digital imaging. Methods: Pattern analysis was performed with 40 adults. A total of 80 images (40 voice onset, 40 voice offset) were analyzed by two speech-language pathologists specialized in voice. Results: The most predominant pattern of voice onset was Type I (nearly simultaneous initiation of vocal folds vibration with the contact of arytenoid cartilages) in all male and female groups, followed by Type II (delayed initiation of vocal folds vibration after contact of arytenoid cartilages) in both groups. Type III voice onset, characterized by the initiation of vocal folds vibration before contact of arytenoid cartilages was only identified in one normal female. On the other hand, Type I, II, and III voice offset patterns were represented in all gender groups. Type I (simultaneously complete cessation of vocal folds vibration with abduction of arytenoid cartilages) was most frequently detected, followed by Type III (abduction of arytenoid cartilages before the cessation of vocal folds vibration), followed by Type II (delayed abduction of arytenoids after complete cessation of vocal folds vibration). However, there was no significant difference between the two gender groups in voice onset and offset patterns (p>.05). Conclusion: Accordingly, the present study may be useful to provide preliminary normative data regarding voice initiation and cessation in Korean young adults.

Voice Problems and Self-Care Practice for Vocal Health: Current Status of Korean Speech-Language Pathologists

Ji Sung Kim(김지성),Seong Hee Choi(최성희) 한국언어청각임상학회 2018 Communication Sciences and Disorders Vol.23 No.2

Objectives: Speech-language pathologists (SLPs) have significant demands on their vocal use and their voice problems have a negative impact on their clinical performance. The purpose of this study was to investigate voice problems and self-care among Korean SLPs. Methods: A total of 187 SLPs completed on online or offline survey and then the data was analyzed. The questionnaire addressed experiences with voice-related problems, knowledge and practice of vocal hygiene, vocal overuse and misuse, and the Voice Handicap Index-10 (VHI-10). Results: About 33% of SLPs had subjective chronic voice problems, about 50% had voice symptoms of vocal fatigue, hoarseness, and hypertension when speaking. Although the majority of SLPs had knowledge regarding vocal hygiene, lower self-evaluation experiences related to voice use and vocal hygiene practice were reported. With respect to the most common reason for voice problems, SLPs answered ‘voice overuse and misuse’ (60.2%) and ‘voice use habits’ (45.4%). SLPs also reported ‘attention and effort’ (55.1%) as the main reason for difficulty with voice hygiene practice. SLPs with subjective voice problems showed significant higher voice symptom (p=.00), overuse and misuse (p=.001), VHI-10 (p=.00), and lower voice hygiene practice (p=.026) than SLPs without subjective voice problems. Voice overuse and misuse and VHI-10 were correlated with voice symptoms (p=.00). Conclusion: The voice problems of SLPs may originate from physical fatigue of the phonatory mechanism. SLPs showed a lack of vocal hygiene and vocal health self-care. A vocal hygiene program tailored to individual SLP’s specific conditions, along with active self-care practice should be established to reduce vocal fatigue and help SLPs achieve vocal health.

문서 편집 접근성 향상을 위한 음성 명령기반 모바일 어플리케이션 개발

박주현,박세아,이무늬,임순범 한국멀티미디어학회 2018 멀티미디어학회논문지 Vol.21 No.11

Voice Command systems are important means of ensuring accessibility to digital devices for use in situations where both hands are not free or for people with disabilities. Interests in services using speech recognition technology have been increasing. In this study, we developed a mobile writing application using voice recognition and voice command technology which helps people create and edit documents easily. This application is characterized by the minimization of the touch on the screen and the writing of memo by voice. We have systematically designed a mode to distinguish voice writing and voice command so that the writing and execution system can be used simultaneously in one voice interface. It provides a shortcut function that can control the cursor by voice, which makes document editing as convenient as possible. This allows people to conveniently access writing applications by voice under both physical and environmental constraints.

음성 사용자 인터페이스(VUI)의 한국어 대우법과 음성 연령에 대한 사용자 경험

이병기,윤재영 한국에이치씨아이학회 2019 한국HCI학회 논문지 Vol.14 No.4

This study investigates the 'likability' and 'trust' of the user through the Korean Honorific method and the voice ages of the interface in the voice user interface. The interface-user relationship is defined as the speaker-listener relationship, In addition, participants were divided into gender and age to see the results. Experimental results showed differences in 'likability' and 'trust' through Korean Honorific's High-Low Way and interface's Yong-Old voice. First, likability is dominated by the low-way ('hae' style), and if the interface voice is Yong-Voice. Conversely, in the case of trust, the Honorific method prevails if it is a High-Way ('hapsyo' + 'haeyo' style), and the interface voice is Old-Voice. Considering the fact that most of the voice user interfaces that are currently being used in the market are based on the heuristic method of using a combination of a hash body and a body and borrowing the voice of young women in their 20s and 30s, We can expect to be able to be based on the design of directionality. 본 연구는 음성을 통해 사용자와 상호작용하는 음성 사용자 인터페이스 (Voice User Interface, VUI) 환경에서 한국어 대우법 방식과 인터페이스의 음성 연령이 사용자에게 미치는 호감도 및 신뢰도를 알아본다. 인터페이스와사용자의 관계를 화자와 청자의 관계로 규정하여 사용자의 성별에 따라, 연령에 따라 나누어 결과를 살펴보았다. 실험연구를 통해 분석 결과, 인터페이스의 대우법 방식(높임-낮춤)과 인터페이스의 음성연령(젊은-나이든)에 대해 사용자가 느끼는 호감도와 신뢰도는 유의한 차이를 보였다. 호감도의 경우 대우법의 낮춤(해체) 방식이 우세했으며, 연령이 20대 음성인 경우 호감도가 우세했다. 반대로 신뢰감에서는 대우법이 높임(합쇼체+해요체)방식일 때와 40대 음성일 경우 우세한 것으로 나타났다. 현재 상용화되고 있는 스피커형태의 음성 사용자 인터페이스가 대부분 합쇼체와해요체를 혼용하는 ‘높임’ 방식의 대우법을 사용하는 점과 20-30대 ‘젊은’ 여성의 음성을 차용하고 있는 점을감안할 때, 본 연구가 다양한 음성 사용자 인터페이스의 방향성을 설계하는데 기초될 수 있을 것으로 기대한다.

Pattern Analysis of Voice Onset and Offset in Normal Adults Using High-Speed Digital Imaging

Seong Hee Choi(최성희),Chi-Sun Oha(오치선),Chul-Hee Choi(최철희) 한국언어청각임상학회 2015 Communication Sciences and Disorders Vol.20 No.4

발성 개시와 종료는 발성 동안 공기역학적, 생체역학적 변화를 포함하며 불규칙적이며 카오스적인 음성 신호를 보인다. 본 예비연구는 초고속 성대촬영 이미지를 이용하여 성별에 따른 발성 개시와 종료 패턴의 특징을 조사하였다. 방법: 40명의 청년층을 대상으로 패턴 분석을 실시하였다. 두 명의 음성언어치료사가 총 80개의 이미지(40개의 발성 개시, 40개의 발성 종료 이미지)를 분석하였다. 결과: 가장 두드러진 발성개시 패턴은 남녀 모든 집단에서 제I형(피열연골의 접촉과 거의 동시에 성대 진동의 시작)이었으며, 그 다음으로는 남녀 모두 제II형(피열연골의 접촉 후 지연된 성대 진동의 시작)의 발성 개시 패턴이었다. 피열연골의 접촉 전 성대 진동이 시작되는 특성을 보이는 제III형의 발성 개시 형태는 여성 한 명에서만 관찰되었다. 한편, 발성 종료 패턴은 모든 성별 집단에서 제I, II, III형이 관찰되었다. 남녀 모두 발성종료 제I형(피열연골의 외전과 거의 동시에 성대 진동의 정지)이 가장 빈번히 관찰되었으며, 그 다음으로는 남녀 모두 제III형(성대 진동의 정지 전 피열연골의 외전, 즉, 피열연골 외전 후 지연된 성대 진동)이었으며, 그 다음으로는 제II형(성대 진동의 완전한 정지 후 지연된 성대의 외전)의 패턴이었다. 그러나, 두 성별 간 발성 개시와 종료 패턴 간에 유의한 차이는 보이지 않았다(p>.05). 논의 및 결론: 따라서, 본 연구는 한국인 정상 청년층의 발성 개시와 종료 패턴에 대한 예비적인 규준 자료를 제공하는 데 유용할 것으로 보인다. Voice onset and offset include dynamic changes of aerodynamic and biomechanical properties showing irregular and chaotic voice signals during phonation. This preliminary study investigated voice onset and offset patterns based on gender using high-speed digital imaging. Methods: Pattern analysis was performed with 40 adults. A total of 80 images (40 voice onset, 40 voice offset) were analyzed by two speech-language pathologists specialized in voice. Results: The most predominant pattern of voice onset was Type I (nearly simultaneous initiation of vocal folds vibration with the contact of arytenoid cartilages) in all male and female groups, followed by Type II (delayed initiation of vocal folds vibration after contact of arytenoid cartilages) in both groups. Type III voice onset, characterized by the initiation of vocal folds vibration before contact of arytenoid cartilages was only identified in one normal female. On the other hand, Type I, II, and III voice offset patterns were represented in all gender groups. Type I (simultaneously complete cessation of vocal folds vibration with abduction of arytenoid cartilages) was most frequently detected, followed by Type III (abduction of arytenoid cartilages before the cessation of vocal folds vibration), followed by Type II (delayed abduction of arytenoids after complete cessation of vocal folds vibration). However, there was no significant difference between the two gender groups in voice onset and offset patterns (p>.05). Conclusion: Accordingly, the present study may be useful to provide preliminary normative data regarding voice initiation and cessation in Korean young adults.

성악가의 음성위생 습관과 한국어판 음성관련 삶의 질(K-VRQOL) 특성

강하늘(Kang, Haneul),김선희(Kim, Seonhee),유재연(Yoo, Jeayeon) 한국음성학회 2018 말소리와 음성과학 Vol.10 No.3

The purpose of this study was to investigate vocal hygiene habits and Korean Voice-Related Quality of Life (K-VRQOL) among classical singers. A total of 128 classical singers filled an online voice and K-VRQOL questionnaire, and the results were analyzed. In order to investigate the characteristics of K-VRQOL according to age groups, and the presence or absence of a history of voice problems, we conducted a two-way ANOVA. The results are as follows. Of the 128 classical singers, 28 (21.9%) with a history of voice problems said that excessive conversation, singing practice, and yelling were the causes of their voice problems. The symptoms of voice problems were fatigue, loss of range, hoarseness, and breathiness. In addition, classical singers were less likely to smoke, or to drink alcohol or caffeine. The K-VRQOL was highly correlated with all sub-domains. There was a statistically significant difference according to age groups (p<.05) and history of voice problems (p<.01). There was no correlation between age groups and history of voice problems. Voice management is important because classical singers can ruin their voice by speaking, and the risk of voice disorder is high. Voice problems affect quality of life. In future studies, it is necessary to obtain information on the subjective voice characteristics of classical singers by examining the relationship between their voice hygiene habits and VHI, SVHI, and K-VRQOL.

보이스오버의 유형과 1960년 전후의 한국영화

장우진(Chang Woo Jin) 한국영화학회 2004 영화연구 Vol.0 No.24

Voice-over is a compound word. It consists of 'voice' and 'over' . 'Voice' means the sound that we can hear when someone is speaking. 'Over' is related to the relationship between the sound and the images; when a voice doesn't come from the images but from another level of the discourse, we call it voice-over. There are two types of voice-over. One is voice-over heard by the subject. The other is voice-over telling. 'Voice-over heard' can be retrospective or hallucinatory. Voice-over telling' is divided into two categories: the voice-over that tells and controls the whole Story and the voice-over that is contained in the story. Voice-over telling and controlling the whole story can come from either outside or inside of the diegesis. If it comes from outside of the diegesis, we call it extra-diegetic. The speaker of the extra-diegetic voice-over can be identified or remain anonymous. If the voice-over comes from inside of the diegesis, we call it intra-diegetic. The listener of the intra-diegetic voice-over can be another character or spectators. When a voice-over is contained in the story, it must be one of the next three: character voice-over contained, interior monologue, or substitution for written narration. Voice-over seems to begin being used around 1955 in Korean films. In 1950's, most of them were contained in the story. It seemed that voice-over began to tell and control the whole story by extra-diegetic narrator around 1960. For example, that kind of voice-over was used in Ah! White Tiger Kim Gu(1960), Concubine Chang(196l), Gu Bong-Sea's Parvenu(196l), etc. Among them, Gu Bong-Sea's Parvenu used diverse types of voice-over most variously. This essay analyses the use and effects of each type of voice-overs in G≪ Bong-Sea's Parvenu. It might be helpful to find out how elaboratory each voice-over was used in Korean films around 1960.

언어 모델 기반 음성 특징 추출을 활용한 생성 음성 탐지

김승민,박소희,최대선,Seung-min Kim,So-hee Park,Dae-seon Choi 한국정보보호학회 2024 정보보호학회논문지 Vol.34 No.3

최근 음성 생성 기술의 급격한 발전으로, 텍스트만으로도 자연스러운 음성 합성이 가능해졌다. 이러한 발전은 타인의 음성을 생성하여 범죄에 이용하는 보이스피싱과 같은 악용 사례를 증가시키는 결과를 낳고 있다. 음성 생성 여부를 탐지하는 모델은 많이 개발되고 있으며, 일반적으로 음성의 특징을 추출하고 이러한 특징을 기반으로 음성 생성 여부를 탐지한다. 본 논문은 생성 음성으로 인한 악용 사례에 대응하기 위해 새로운 음성 특징 추출 모델을 제안한다. 오디오를 입력으로 받는 딥러닝 기반 오디오 코덱 모델과 사전 학습된 자연어 처리 모델인 BERT를 사용하여 새로운 음성 특징 추출 모델을 제안하였다. 본 논문이 제안한 음성 특징 추출 모델이 음성 탐지에 적합한지 확인하기 위해 추출된 특징을 활용하여 4가지 생성 음성 탐지 모델을 만들어 성능평가를 진행하였다. 성능 비교를 위해 기존 논문에서 제안한 Deepfeature 기반의 음성 탐지 모델 3개와 그 외 모델과 정확도 및 EER을 비교하였다. 제안한 모델은 88.08%로 기존 모델보다 높은 정확도와 11.79%의 낮은 EER을 보였다. 이를 통해 본 논문에서 제안한 음성 특징 추출 방법이 생성 음성과 실제 음성을 판별하는 효과적인 도구로 사용될 수 있음을 확인하였다. Recent rapid advancements in voice generation technology have enabled the natural synthesis of voices using text alone. However, this progress has led to an increase in malicious activities, such as voice phishing (voishing), where generated voices are exploited for criminal purposes. Numerous models have been developed to detect the presence of synthesized voices, typically by extracting features from the voice and using these features to determine the likelihood of voice generation.This paper proposes a new model for extracting voice features to address misuse cases arising from generated voices. It utilizes a deep learning-based audio codec model and the pre-trained natural language processing model BERT to extract novel voice features. To assess the suitability of the proposed voice feature extraction model for voice detection, four generated voice detection models were created using the extracted features, and performance evaluations were conducted. For performance comparison, three voice detection models based on Deepfeature proposed in previous studies were evaluated against other models in terms of accuracy and EER. The model proposed in this paper achieved an accuracy of 88.08%and a low EER of 11.79%, outperforming the existing models. These results confirm that the voice feature extraction method introduced in this paper can be an effective tool for distinguishing between generated and real voices.

목소리에 주목한 소설 읽기 교육의 이론적 고찰

우수연 ( Woo Sooyeon ) 서울대학교 국어교육과 2024 선청어문 Vol.53 No.-

The purpose of this study is to theorize the concept of ‘voice’, which is distinguished from existing point-of-view theory on novel, by focusing on the utterance of the speakers and characters in the novel. Also this study aims to reveal the method of reading novels and their educational effects through the concept of ‘voice’. The consciousness of the problem starts from the premise that the novel text is ‘written(characterized) speech’ and that the speaker aims at the receiver who listens to his or her speech and tells it through a speech strategy. The existing novel education method of understanding the speaker as a mediated speech has been made in terms of visual aspects such as ‘viewpoint’ and ‘focalization’. It is difficult to shed light on the reader's activity of grasping and constructing the multi-layered voice concept of the novel. Through the fact that the sense of hearing speaker's voice from novel is not actually heard by the reader but become the object of the reader's imagination and implementation through the text, the sense of hearing the voice layers the semantic system of the narrative text and maximizes the reader's role. The speaker in the novel chooses a specific speech strategy for the receiver to listen to his or her speech. This study embodies a novel speech structure diagram that focuses on the voice based on the four components of the narrator, narratee, character, and the implied listener. According to the speech act theory, speech itself has performance and performance speech has not only meaning but also its effect. In the theory of verbal act that takes place in the interaction between the speaker and the listener, it is necessary to clarify the role of the listener in grasping the utterance behavior and its effect in the novel text. Therefore, this study calls the reader as a listener as the ‘auscultator’, who listens to the utterance of the novel text and understands the utterance strategy and effectiveness. The auscultator recognizes that speech is the product of the speaker's intention and strategy in the reproduction process, leads to his interpretation of the novel, and further critically recognizes and reconstructs speech concealed in the reproduction process. Based on the above, reading a novel focused on the voice can be defined as reading a novel by an auscultator who pays attention to the speaker's utterance in the novel. And as a auscultator's role, it consists of three principles : ‘understanding narrator’s utterance strategy by focusing on words’, ‘discriminating voice by focalizing on voice’ and ‘possessing text by reconstructing the voice’. Reading novels by focusing on voice is meaningful in that it newly conceptualized the learning reader as a ‘reader as a listener’. This contains possibility of embodying the characteristics of novel reading, which is a cognitive and imagination, through the performance of the reader as a listener in novel reading. Furthermore, this study is meaningful in that it illuminates the existing novel communication composition, which was composed in a single and monotonous way, through a new sense of hearing, and that ‘polyphony’, the essence of novel text, is embodied as a specific novel reading method and applied to actual text interpretation.

실재적 대상으로서의 목소리

김용수(Yongsoo Kim) 한국비평이론학회 2016 비평과이론 Vol.21 No.2

정신분석에서 목소리는 여전히 미지의 영역으로 남아있다. 그것은 지금까지 청각 욕동(invocatory drive)의 대상, 의미화 작용의 잉여 부산물 정도로만 알려져 왔을 뿐 시각(scopic) 욕동의 대상인 응시(gaze)에 비해 알려진 바가 별로 없었다. 이런 의미에서 최근 영어로 번역되어 출판된 라캉(Jacques Lacan)의 세미나 10권: 불안(Seminar X: Anxiety)은 충분하지는 않지만 어느 정도 갈증을 해소해줄 수 있는 중요한 텍스트이다. 이 세미나에서 라캉은 욕동의 다섯 단계에 해당하는 대상 소타자(objet petit a)들 - 젖가슴(breast), 대변(faeces), 남근(phallus), 응시, 목소리(voice) - 을 하나하나 설명해나가면서 목소리에 대해서도 의미심장한 대목들을 여럿 남긴다. 필자는 목소리가 지닌 대상 소타자로서의 일반적 특징과 특수성을 제대로 파악하기 위해 각 단계별로 대상들을 비교 검토한 뒤에 대상 목소리와 관련한 라캉의 설명들을 정리하면서 본격적인 목소리의 이론을 구성하고자 한다. 목소리는 청각 욕동의 실재적 대상으로서 의미를 벗어난 비명이다. 그것은 신화적으로 원초적 아버지의 끔찍한 비명으로 표출되고, 이로 인한 불안을 해소하는 과정에서 비명은 초자아적 명령으로 전환된다. 세미나 10권은 원초적 아버지라는 신화적 맥락 속에서 비명의 무의미한 목소리가 초자아의 구체적인 명령으로 재탄생하는 기제를 밝혀준다. 나아가서 라캉이 간과하는 비명의 음악적 가능성은 금지의 한계를 넘어선 여성적 희열의 목소리를 가리킨다. 오페라의 역사에서 끊임없이 추구된 ‘천사의 비명’은 소리의 강렬함으로 희열을 표현하고, 베르니니(Gian Lorenzo Bernini)의 조각상은 소리 없는 형상으로 희열의 목소리를 구현한다. 오랫동안 목소리는 응시의 그늘에 가려져 있었다. 이제 목소리가 품은 무한한 가능성에 귀 기울일 때다. Voice remains a relatively unknown object in Psychoanalysis. Despite its theoretical and practical implications, not much is known about the object voice, except that it is the object of the invocatory drive and the surplus remainder of signification. Jacques Lacan’s Seminar X: Anxiety, in this regard, which was recently published in English, is a significant text that will illuminate the auditory dimension of the objet petit a. As explaining the five different objects - breast, faeces, phallus, gaze, and voice - that correspond to the five levels of the drive, Lacan elaborates the main characteristics of the object voice in a revealing way. This paper constructs a psychoanalytic theory of voice by examining its general and particular features in relation to other objects. Voice as a real object is ultimately a meaningless scream. It is mythically expressed as the appalling cry of the primal father; in resolving the anxiety engendered by the intolerable roar, the subjects turn it into the superegoic commands. Seminar X provides a valuable insight into how the meaningless voice of the cry is transformed into the specific commands of the superego in the context of the primal horde. Moreover, the musical possibility of the scream Lacan happens to overlook points to the voice of feminine jouissance beyond the limits of prohibition. “The angel’s cry” that has been passionately pursued in the entire operatic history expresses the feminine jouissance with a sounding intensity; Bernini’s sculpture embodies the voice of female ecstasy in a soundless form. Voice has long been shadowed by the gaze. Now is the time to listen to the unlimited possibilities of the object voice.