• Smart Structures and Systems
• /
• 제29권3호
• /
• pp.475-484
• /
• 2022

The loss of cable tension for civil infrastructure reduces structural bearing capacity and causes harmful deformation of structures. Currently, most of the structural health monitoring (SHM) approaches for cables rely on contact transducers. This paper proposes a cable tension identification technology using percussion sound, which provides a fast determination of steel cable tension without physical contact between cables and sensors. Notably, inspired by the concept of tensioning strings for piano tuning, this proposed technology predicts cable tension value by deep learning assisted classification of "percussion" sound from tapping a steel cable. To simulate the non-linear mapping of human ears to sound and to better quantify the minor changes in the high-frequency bands of the sound spectrum generated by percussions, Mel-frequency cepstral coefficients (MFCCs) were extracted as acoustic features to train the deep learning network. A convolutional neural network (CNN) with four convolutional layers and two global pooling layers was employed to identify the cable tension in a certain designed range. Moreover, theoretical and finite element methods (FEM) were conducted to prove the feasibility of the proposed technology. Finally, the identification performance of the proposed technology was experimentally investigated. Overall, results show that the proposed percussion-based technology has great potentials for estimating cable tension for in-situ structural safety assessment.

• 음성과학
• /
• 제10권2호
• /
• pp.7-25
• /
• 2003

Predicting the right prosodic elements is a key factor in improving the quality of synthesized speech. Prosodic elements include break, pitch, duration and loudness. Pitch, which is realized by Fundamental Frequency (F0), is the most important element relating to the quality of the synthesized speech. However, the previous method for predicting the F0 appears to reveal some problems. If voiced and unvoiced sounds are not correctly classified, it results in wrong prediction of pitch, wrong unit of triphone in synthesizing the voiced and unvoiced sounds, and the sound of click or vibration. This kind of feature is usual in the case of the transformation from the voiced sound to the unvoiced sound or from the unvoiced sound to the voiced sound. Such problem is not resolved by the method of grammar, and it much influences the synthesized sound. Therefore, to steadily acquire the correct value of pitch, in this paper we propose a new model for predicting and classifying the voiced and unvoiced sounds using the CART tool.

• 한국소음진동공학회:학술대회논문집
• /
• 한국소음진동공학회 2008년도 춘계학술대회논문집
• /
• pp.795-800
• /
• 2008

The sound transmission loss is a key index to evaluate the acoustic performance of a car-body structure at the high frequency range. From this paper, a new validation method for in-situ sound transmission loss is proposed. First, in-situ sound transmission loss is measured by using PU intensity probe on the condition of complete vehicle. Second, validation test, which is consisted of internal, external and total frequency response function test, is performed by using volume acceleration source and microphones. Then, these test results are compared to validate the accuracy of in-situ sound transmission loss. Finally, the test result of in-site method is compared with results of two reverberant room test method and SEA analytical method. The reliability of in-situ method is confirmed by these procedures.

• 한국소음진동공학회:학술대회논문집
• /
• 한국소음진동공학회 2006년도 추계학술대회논문집
• /
• pp.946-949
• /
• 2006

This study is a trial to make a design chart of sound insulation for multiple panel. Dilatational frequency, ${\Large f}_d$ becomes a key factor for optimal design since it acts like a turning point in sound insulation performance of such panels. Hence, in tuning the ${\Large f}_d$ optimally, elastic modulus of core material and thickness of the skin panel is designated to parameters. Based on these parameter, a design chart of sound insulation for multiple panel is made. Its applicability is proved by the case study of High noise reduction panel.

• 한국음향학회지
• /
• 제39권5호
• /
• pp.424-434
• /
• 2020

본 논문은 콘텐츠 플랫폼에서 절대음량을 산출하는 알고리즘을 서술한다. 절대음량은 개별 음원을 대표하는 하나의 음량이며, 실용적인 범위 내에서, 음원파일 자체의 디지털 영역과 스피커로 출력되는 물리 영역의 음량단위를 통합하여 활용할 수 있도록 설계한 개념이다. 이 개념이 콘텐츠 플랫폼 등에서 실제로 활용되기 위해서는 마스터링 엔지니어의 청음을 거치지 않아도, 자동으로 절대음량 수치를 도출할 수 있어야 한다. 사람이 개별 단일 음원의 대표 음량을 인지하는 핵심 파라미터는 "주파수, 최대에너지, 에너지 변동 계수, 인지가중 에너지 분포도" 영역이며, 가중치 정규화를 통하여 절대음량 수치를 산출하였다.

• 디지털콘텐츠학회 논문지
• /
• 제9권2호
• /
• pp.331-339
• /
• 2008

본 논문에서는 3D 립싱크 애니메이션에 필요한 키 데이터를 생성하는 효율적이고 정확한 시스템 개발을 제안한다. 여기서 개발한 시스템은 한국어를 기반으로 발화된 음성 데이터와 텍스트 정보에서 한국어 음소를 추출하고 분할된 음소들을 사용하여 정확하고 자연스러운 입술 애니메이션 키 데이터를 계산한다. 이 애니메이션 키 데이터는 본 본문에서 개발한 3D 립싱크 애니메이션 시스템뿐만 아니라 상업적인 3D 얼굴 애니메이션 시스템에서도 사용된다. 전통적인 3D 립싱크 애니메이션 시스템은 음성 데이터를 영어 음소 기반으로 음소를 분할하고 분할된 음소를 사용하여 립싱크 애니메이션 키 데이터를 생성한다. 이러한 방법의 단점은 한국어 콘텐츠에 대해 부자연스러운 애니메이션을 생성하고 이에 따른 추가적인 수작업이 필요하다는 것이다. 본 논문에서는 음성 데이터와 텍스트 정보에서 한국어 음소를 추출하고 분할된 음소를 사용하여 자연스러운 립싱크 애니메이션을 생성하는 3D 립싱크 애니메이션 시스템을 제안한다.

• 동의생리병리학회지
• /
• 제16권6호
• /
• pp.1106-1109
• /
• 2002

The analysis on the therapic music of chinese five-sounds studied in order to the stand ardization of the five-sounds. The results were as follows; The central sound of Gung(宮) becomes Do. This music goes on slow speed from first to last and appears the bright, quiet and stable mood. The central sound of Sang(商) becomes Re. This music mainly uses the brass and the fist appears the heavy feeling but gradually becomes ascending mood. The central sound of Kang(角) becomes Mi. This music mainly uses the wood-wind imstrument and appears the most light and vivid feeling. The central sound of Chi(徵) becomes Sol. All instruments sound grandly in the first part, and subsequently sound gentle. The central sound of Woo(羽) becomes La. This music appears dark and tragic mood that is generally used in the minor key. It is played slowly on the whole and appears lyrical feeling in the process of controlling and quiet. In the future, the study on diagnosis and therapy through the standardization of the five-sounds will be continued.

• 한국소음진동공학회:학술대회논문집
• /
• 한국소음진동공학회 2004년도 추계학술대회논문집
• /
• pp.857-861
• /
• 2004

An active method that improves low frequency characteristics of a reverberation chamber has been proposed. Spatial uniformity of sound pressure at low frequency has been increased by applying active noise control system. The system acts as frequency selective damping materials, which decreases resonance at low frequency and leave high frequency sound field as it is. Linearity of the chamber, which is essential for sound power measurement in the chamber, is guaranteed since fixed control logic has been adopted. Simple application and tests have been done in a small sized rectangular reverberation chamber which has 400Hz cutoff, Loud speakers were used to simulate a noise source and to construct an active noise control system. The spatial distribution of a sound field at 250Hz, 315Hz and 400Hz bands has been measured before and after control. The standard deviation of sound field has been decreased from 3.4dB to 2.5dB. The results show the possibility of active control system as a sound diffuser.

• 한국소음진동공학회논문집
• /
• 제20권11호
• /
• pp.1018-1024
• /
• 2010

The printer noise consists of the noise of the various components and parts such as motor, fan and solenoid. And the human's printing sound recognition shows various aspects when the printer starts to print papers because the components operate at the same time. Especially, printers are usually installed in the quiet office room. Therefore the printing noise is related to its competitiveness in the market. The importance of the printer sound qualities is increasing and it is necessary to develop the sound quality evaluation system, so it is a key point to identify the noise source of the printer and develop the sound quality index to each component. By using this evaluation system, it is possible to evaluate the sound quality of a prototype printer compared to the already existing one. In this paper, the printer sound quality evaluation system was developed by the following steps. Firstly, the signal processing method was applied to the recorded printing sound to identity and split the noise of components. Secondly, the MLR(multiple linear regression) method and the psychoacoustics were used to develop the sound quality index. Finally, the improvement of the printer sound quality is possible by using the result of the MLR and the path analysis. The output of this research will be applied to the development of a new printer.

• 한국소음진동공학회:학술대회논문집
• /
• 한국소음진동공학회 2012년도 추계학술대회 논문집
• /
• pp.630-635
• /
• 2012

Nearfield acoustic holography method predicts an unmeasured sound field, therefore it depends on its prediction methods. In particular, if one has radiators or scatters, which cannot be expressed by simple geometry, then inverse boundary element method (BEM) is normally employed to reconstruct the sound field induced by sound sources with irregular profiles. The characteristics of boundary element, including the element shape, characteristic length, order of shape function and others, affect the reconstruction error. Investigating the errors by means of changing these factors will provide a guide line for selecting appropriate factors, associated with the elements of BEM. These factors are investigated by numerical simulations, and the accuracies with respect to the variant factors are compared. Novel suggestions for selecting appropriate boundary element factors are described based on the simulation results.