현악기의 물리적 모델링을 위한 최적의 멀티코어 프로세서 아키텍처 탐색

강명수,최지원,김용민,김종면,Kang, Myeong-Su,Choi, Ji-Won,Kim, Yong-Min,Kim, Jong-Myon 한국음향학회 2011 韓國音響學會誌 Vol.30 No.5

물리적 모델링 기반 음 합성 알고리즘은 음 합성 시 많은 연산량을 요구하며 이는 실시간 음 합성을 저해한다. 이러한 문제를 해결하기 위해 본 논문에서는 물리적 모델링 기반 현악기 사운드 엔진을 멀티코어 프로세서에 구현하고, 사운드 엔진을 위한 최적의 멀티코어 프로세서 구조를 제안한다. 대상 현악기의 단위음을 합성하기 위해 각 프로세싱 엘리먼트 (processing element, PE)당 합성하는 샘플 (sample-per-processing element, SPE) 수를 변화시키는 실험을 통해 시스템의 성능 (system performance), 시스템 면적 효율 (area efficiency), 에너지 효율 (energy efficiency)을 각각 측정하고, 측정된 결과를 바탕으로 최적의 멀티코어 프로세서 구조를 선택하였다. 모의실험 결과, 어쿠스틱 기타는 SPE가 5,513과 2,756일 때 가장 높은 시스템 면적 효율과 에너지 효율을 보였으며, 클래식 기타는 SPE가 22,050과 5,513일 때 시스템 면적 효율과 에너지 효율이 가장 높았다. 또한 이를 이용하여 44.1 kHz의 샘플링율을 갖도록 대상 악기의 단위음을 합성한 결과 원음과 스펙트럼에서 매우 유사함을 확인할 수 있었고, 울산대학교 대학원생 및 교수 10명을 대상으로 실시한 MUSHRA 주관 청취 테스트에서도 좋은 결과를 얻었다. Physics-based sound synthesis usually requires high computational costs and this results in a restriction of its use in real-time applications. This motivates us to implement the sound synthesis algorithm of plucked-string instruments using multi-core processor architectures and determine the optimal processing element (PE) configuration for the target instruments. To determine the optimal PE configuration, we evaluate the impacts of a sample-per-processing element (SPE) ratio that is defined as the amount of sample data directly mapped to each PE on system performance and both area and energy efficiencies using architectural and workload simulations. For the acoustic guitar, the highest area and energy efficiencies are achieved at a SPE ratio of 5,513 and 2,756, respectively, for the synthesis of musical sounds sampled at 44.1 kHz. In the case of the classical guitar, the maximum area and energy efficiencies are achieved at a SPE ratio of 22,050 and 5,513, respectively. In addition, the synthetic sounds were very similar to original sounds in their spectra. Furthermore, we conducted MUSHRA subjective listening test with ten subjects including nine graduate students and one professor from the University of Ulsan, and the evaluation of the synthetic sounds was excellent.

음질 열화를 줄이고 공격에 강인한 오디오 워터마킹 알고리듬

강명수,조상진,정의필,Kang, Myeong-Su,Cho, Sang-Jin,Chong, Ui-Pil 한국음향학회 2009 韓國音響學會誌 Vol.28 No.7

본 논문에서는 오디오 신호의 저작권 보호와 효과적인 음질 개선을 위한 새로운 워터마킹 알고리듬을 제안한다. 제안한 방법은 원 신호에 푸리에 변환을 하여 주파수 영역으로 변환하고 n개의 서브밴드로 균등 분할한다. 각 밴드별 에너지를 계산하여 에너지가 큰 것부터 k개를 선택하고 해당 밴드에서 p개의 주요 피크 성분을 검출하여 길이 m의 워터마크를 삽입한다. 워터마크된 오디오 신호를 청자에게 들려주었을 때 워터마크 삽입으로 인한 오디오 신호의 왜곡을 느끼지 못하였다. 또한, 제안한 방법은 Cox 방법만큼 MP3 압축, 잘라내기 (cropping),주파수 변환 (FFT), 반향 (echo)과 같은 워터마크 공격에 강인하였고 신호 대 잡음비 측면에서는 10 dB이상 우수함을 실험을 통해 확인할 수 있었다. This paper proposes a robust audio watermarking algorithm for copyright protection and improvement of sound quality after embedding a watermark into an original sound. The proposed method computes the FFT (fast Fourier transform) of the original sound signal and divides the spectrum into a subbands. Then, it is necessary to calculate the energy of each subband and sort n subbands in descending order corresponding to its power. After calculating the energy we choose k subbands in sorted order and find p peaks in each selected subbands, and then embed a length m watermark around the p peaks. When the listeners hear the watermarked sound, they do not recognize any distortions. Furthermore, the proposed method is robust as much as Cox's method to MP3 compression, cropping, FFT echo attacks. In addition to this, the experimental results show that the proposed method is generally 10 dB higher than Cox's method in SNR (signal-to-noise ratio) aspect.

유도 전동기의 고장 검출 및 분류를 위한 특징 벡터 추출과 분류기의 다양한 설정에 따른 분류 성능 비교

강명수,뉘엔 투 낙,김용민,김철홍,김종면,Kang, Myeong-Su,Nguyen, Thu-Ngoc,Kim, Yong-Min,Kim, Cheol-Hong,Kim, Jong-Myon 한국음향학회 2011 韓國音響學會誌 Vol.30 No.8

최근 항공 산업, 자동차 산업 등의 산업 현장에서 유도 전동기의 사용이 증대되고 있으며, 유도 전동기는 산업 현장에서 중요한 역할을 하고 있다. 따라서 유도 전동기의 고장으로 인한 피해를 최소화하기 위해 유도 전동기의 고장 검출 및 분류 시스템의 개발이 중요한 문제로 대두되고 있다. 이와 같은 이유로 본 논문에서는 유도 전동기의 고장을 조기에 검출하고 진단하기 위해 에너지 (short-time energy)와 특이치 분해와 이산 코사인 변환과 특이치 분해를 이용한 특징 벡터 추출 방법을 제안하였고, 이를 역 전파 신경 회로망과 다층 서포트 벡터 머신의 입력으로 이용하여 유도 전동기의 고장을 유형별로 분류하였다. 하지만 본 논문에서는 역 전파 신경 회로망과 다층 서포트 벡터 머신을 분류기로 사용함에 있어 역 전파 신경 회로망은 신경망을 구성하는 입력 뉴런 수, 은닉 뉴런 수, 학습 알고리즘에 의해 분류 성능이 달라지며, 다층 서포트 벡터 머신은 커널 함수로 사용한 가우시안 방사 기저 함수의 표준 편차 값에 따라 분류 성능이 달라지는 점을 고려하여 여러 가지 조건하에서의 실험을 통해 높은 분류 성능을 보이는 설정 방법을 제시하였다. The use of induction motors has been recently increasing with automation in aeronautical and automotive industries, and it playes a significant role. This has motivated that many researchers have studied on developing fault detection and classification systems of an induction motor in order to minimize economical damage caused by its fault. With this reason, this paper proposed feature vector extraction methods based on STE (short-time energy)+SVD (singular value decomposition) and DCT (discrete cosine transform)+SVD techniques to early detect and diagnose faults of induction motors, and classified faults of an induction motor into different types of them by using extracted features as inputs of BPNN (back propagation neural network) and multi-layer SVM (support vector machine). When BPNN and multi-lay SVM are used as classifiers for fault classification, there are many settings that affect classification performance: the number of input layers, the number of hidden layers and learning algorithms for BPNN, and standard deviation values of Gaussian radial basis function for multi-layer SVM. Therefore, this paper quantitatively simulated to find appropriate settings for those classifiers yielding higher classification performance than others.

물리적 모델링 합성법에 기반을 둔 줄 없는 기타 구현

강명수,조상진,정의필,Kang, Myeong-Su,Cho, Sang-Jin,Chong, Ui-Pil 한국음향학회 2009 韓國音響學會誌 Vol.28 No.2

This paper describes the non-stringed guitar composed of laser strings, frets, sound synthesis algorithm and a processor. The laser strings that can depict stroke and playing arpeggios comprise laser modules and photo diodes. Frets are implemented by voltage divider. The guitar body does not need to implement physically because commuted waveguide synthesis is used. The proposed frets enable; players to represent all of chords by the chord glove as well as guitar solo. Sliding, hammering-on and pulling-off sounds are synthesized by using parameters from the voltage divider. Because the pitch shifting corresponds to the time-varying propagation speed in the digital waveguide model, the proposed model can synthesize vibrato as well. After transformation of signals from the laser strings and frets into parameters for synthesis algorithm, the digital signal processor, TMS320F2812, performs the real-time synthesis algorithm and communicates with the DAC. The demonstration movieclip available via the Internet shows one to play a song, 'Arirang', synthesized by proposed algorithm and interfaces in real-time. Consequently, we can conclude that the proposed synthesis algorithm is efficient in guitar solo and there is no problem to play the non-stringed guitar in real-time. 본 논문에서 제안한 줄 없는 기타는 크게 레이저 현과 프렛, 음 합성 알고리듬과 프로세서로 구성된다. 레이저 현은 레이저 모듈과 포토다이오드를 이용하여 스트로크와 아르페지오를 표현할 수 있도록 하였고, 프렛은 전압 분배기를 이용하여 구현하였다. 몸통은 물리적 모델링 합성법을 이용하였기 때문에 줄 없는 기차에는 울림통의 역학을 하는 물리적인 몸통이 없다. 제안한 기타의 프렛은 실제 프렛을 동일하게 표현할 수 있어 기존의 코드 글러브를 이용한 로드 표현뿐 아니라 실제 기타와 같은 솔로 연주도 가능하다. 해머링 온, 풀링 오프, 슬라이딩과 같이 프렛 변화가 있는 연주음은 전압 분배기로부터 받은 프렛의 정보를 파라미터로 사용하여 합성한다. 연구법에 따른 음의 피치 변화는 디지털 도파관 모델에서 파동의 전파 속도 변화로 표현하였다. 이 합성 모델은 동일 프렛에서 현의 장력을 변화시켜 연주하는 비브라토 음도 합성 할 수 있다. 레이저 현과 프렛으로부터 받아들인 정보를 합성 알고리듬의 파라미터로 변환하여 기타 음을 생성하고 이를 실시간으로 출력할 수 있도록 TMS320F2812를 사용하였다. 웹에 공개한 동영상에는 제안한 알고리듬과 인터페이스를 이용하여 실시간으로 합성한 '아리랑' 연주를 볼 수 있다. 제안한 알고리듬이 피치 변화를 표현하는 기타 솔로 연주법에 효과적이고 줄 없는 기타로 실시간 연주가 가능함을 확인할 수 있다.

Audio Segmentation and Classification Using Support Vector Machine and Fuzzy C-Means Clustering Techniques

강명수,김철홍,김종면,Nguyen, Ngoc,Kang, Myeong-Su,Kim, Cheol-Hong,Kim, Jong-Myon 한국정보처리학회 2012 정보처리학회논문지B Vol.19 No.1

최근 멀티미디어 정보가 급증함에 따라 콘텐츠 관리에 대한 요구도 함께 증가되고 있다. 이에 오디오 분할 및 분류는 멀티미디어 콘텐츠를 효과적으로 관리할 수 있는 대안이 될 수 있다. 따라서 본 논문에서는 동영상에서 취득한 오디오 신호를 분할하고, 분할된 오디오 신호를 음악, 음성, 배경 음악이 포함된 음성, 잡음이 포함된 음성, 묵음(silence)으로 분류하는 정확도가 높은 오디오 분할 및 분류 알고리즘을 제안한다. 제안하는 알고리즘은 오디오 분할을 위해 서포트 벡터 머신(support vector machine, SVM)을 이용하였다. 오디오 신호의 분류를 위해서는 분할된 오디오 신호의 특징을 추출하고 이를 퍼지 클러스터링 알고리즘(fuzzy c-means, FCM)의 입력으로 사용하여 각 계층으로 오디오 신호를 분류하였다. 제안하는 알고리즘의 평가는 분할과 분류에 대해 각각 그 성능을 평가하였으며, 분할 성능 평가는 정확도율(precesion rate)과 오차율(recall rate)을 이용하였으며, 분류 성능 평가는 정확성(classification accuracy)을 사용하였다. 또한 오디오 분할의 경우는 이진 분류기와 퍼지 클러스터링을 이용한 기존의 알고리즘과 그 성능을 비교하였다. 모의 실험 결과, 제안한 알고리즘의 분류 성능이 기존 알고리즘 보다 정확도율과 오차율 면에서 모두 우수하였다. The rapid increase of information imposes new demands of content management. The purpose of automatic audio segmentation and classification is to meet the rising need for efficient content management. With this reason, this paper proposes a high-accuracy algorithm that segments audio signals and classifies them into different classes such as speech, music, silence, and environment sounds. The proposed algorithm utilizes support vector machine (SVM) to detect audio-cuts, which are boundaries between different kinds of sounds using the parameter sequence. We then extract feature vectors that are composed of statistical data and they are used as an input of fuzzy c-means (FCM) classifier to partition audio-segments into different classes. To evaluate segmentation and classification performance of the proposed SVM-FCM based algorithm, we consider precision and recall rates for segmentation and classification accuracy for classification. Furthermore, we compare the proposed algorithm with other methods including binary and FCM classifiers in terms of segmentation performance. Experimental results show that the proposed algorithm outperforms other methods in both precision and recall rates.

스펙트럼영역 빛간섭단층촬영기로 측정한 한국인 정상안의 황반부 용적

강명수,경성은,장무환,Myeong Su Kang,Sung Eun Kyung,Moo Hwan Chang 대한안과학회 2010 대한안과학회지 Vol.51 No.8

Purpose: To evaluate the mean macular volume in normal Korean eyes using spectral-domain optical coherence tomography. Methods: The present study consisted of 132 patients (212 eyes) with no ophthalmic evidence of retinopathy and who had a best corrected visual acuity of 1.0 or better. The total macular volume was measured using spectral-domain optical coherence tomography and was analyzed according to age group, sex, degree of refractive error and presence of systemic disease such as diabetes and hypertension. Results: The mean total macular volume of all subjects was 10.07 ± 0.45 mm3, with means of 10.13 ± 0.40 mm3, 10.05 ± 0.43 mm3 and 9.97 ± 0.58 mm3 measured for the respective A, B, and C age groups. There was a significant difference between male and female patients. In addition, there was a significant difference between diabetic patients in group C and normal subjects. However, there was no significant difference according to degree of refractive error. Conclusions: The measured value of mean macular volume in normal Korean eyes can be expected to provide a standard value for diagnosing retinal disease and the need for careful follow-up. J Korean Ophthalmol Soc 2010;51(8):1077-1083

안과 장애환자의 임상양상

강명수,경성은,Myeong Su Kang,Sung Eun Kyung 대한안과학회 2010 대한안과학회지 Vol.51 No.1

Purpose: To describe the characteristics of the disabled patients visiting the eye clinic in our institute. Methods: We carried out a retrospective analysis of 35 cases in our clinic from April 2004 to June 2008 using the McBride disability evaluation. We investigated the clinical features and the causes of disorders through visual acuity, visual field and ocular motility. Results: Thirty-three (94.3%) of the 35 patients had disabilities due to trauma; twelve (34.3%) of them were caused by traffic accidents, and 21 (60%) of them were due to blows or lacerations. Other causes of disability were glaucoma and retinal break (5.8%). Nine patients (25.7%) had abnormal findings in the visual field examination, and nine other patients (25.7%) had limitations in ocular motility. Twenty-eight patients (80%) had decreased visual acuity, and nine (25.7%) had multiple symptoms. Conclusions: Considering the contribution of disability estimation of visual field and ocular motility in McBride disability evaluations, we considered the importance of repetitive examinations and evaluations. When patients complained of unexplained decreased visual acuity with no anatomical abnormalities, multifocal ERG and multifocal VEP should be considered in order to distinguish it from malingering or functional visual loss. J Korean Ophthalmol Soc 2010;51(1):95-105

범용 디지털 신호처리기를 이용한 국악기 사운드 엔진 개발

강명수,조상진,권순덕,정의필,Kang, Myeong-Su,Cho, Sang-Jin,Kwon, Sun-Deok,Chong, Ui-Pil 한국음향학회 2009 韓國音響學會誌 Vol.28 No.3

본 논문에서는 TMS3320F2812 신호처리기를 이용하여 가야금과 태평소의 사운드 엔진을 구현하였다. Commuted Waveguide Synthesis (CWS) 기반의 가야금과 태평소 모델을 신호처리기에 탑재하고 악기 선택 버튼을 두어 해당 악기의 사운드 샘플을 매 일정 시간마다 합성하도록 하였다. 합성음은 SPI 통신을 이용하여 DAC로 전송되며 오디오 인터페이스를 거쳐 스피커를 통해 재생된다. 합성 모델의 지연 라인은 합성음의 피치를 조절하는데, 이 지연라인의 길이를 결정하기 위해 GPIO를 이용하여 한 샘플을 합성하는데 필요한 시간을 측정하였다. 가야금은 $28.6{\mu}s$, 태평소는 $21{\mu}s$가 소요되었다. 태평소와 가야금의 동시 발음수를 고려하였을 때 태평소는 동시 발음수 1을 가지므로 $21{\mu}s$, 가야금은 일반적으로 동시 발음수가 2이므로 $57.2{\mu}s$의 연산시간이 필요하다. 이는 실시간 연주가 충분히 가능한시간이다. 제안한 사운드 엔진의 경우, 인터럽트 서비스 루틴에서 각 사운드 샘플의 합성과 DAC로의 전송이 일어난다. 인터럽트 서비스 루틴은 시스템의 안정성을 보장하기 위해 타이머의 주기 매칭 이벤트를 이용하여 $60{\mu}s$마다 주기적으로 호출된다. 이와 같이 합성된 음을 녹음하여 원음과 스펙트럼으로 비교한 결과, 가야금은 원음과 매우 유사한 음을 합성할 수 있었고, 태평소는 '무(無), 황(黃), 태(太), 중(仲)' 음을 제외한 나머지 음에 대해서 태평소의 음색을 잘 표현하는 음을 합성 할 수 있었다. This paper describes a sound engine of Korean traditional instruments, which are the Gayageum and Taepyeongso, by using a TMS320F2812. The Gayageum and Taepyeongso models based on commuted waveguide synthesis (CWS) are required to synthesize each sound. There is an instrument selection button to choose one of instruments in the proposed sound engine, and thus a corresponding sound is produced by the relative model at every certain time. Every synthesized sound sample is transmitted to a DAC (TLV5638) using SPI communication, and it is played through a speaker via an audio interface. The length of the delay line determines a fundamental frequency of a desired sound. In order to determine the length of the delay line, it is needed that the time for synthesizing a sound sample should be checked by using a GPIO. It takes $28.6{\mu}s$ for the Gayageum and $21{\mu}s$ for the Taepyeongso, respectively. It happens that each sound sample is synthesized and transferred to the DAC in an interrupt service routine (ISR) of the proposed sound engine. A timer of the TMS320F2812 has four events for generating interrupts. In this paper, the interrupt is happened by using the period matching event of it, and the ISR is called whenever the interrupt happens, $60{\mu}s$. Compared to original sounds with their spectra, the results are good enough to represent timbres of instruments except 'Mu, Hwang, Tae, Joong' of the Taepyeongso. Moreover, only one sound is produced when playing the Taepyeongso and it takes $21{\mu}s$ for the real-time playing. In the case of the Gayageum, players usually use their two fingers (thumb and middle finger or thumb and index finger), so it takes $57.2{\mu}s$ for the real-time playing.

「한국조리학회지」 수록논문의 연구동향 분석

강명수(Myeong Su Kang) 한국조리학회 2014 한국조리학회지 Vol.20 No.1

This research has been conducted to provide the fundamental data for foodservice & culinary research by analyzing 1,054 papers in ``The Korean Journal of Culinary Research`` from 1995 through 2012 regarding year, theme, material, and subject. The released papers are 58.6 a year and almost 50 percent of related papers have been published, considering that 511 papers have been released for the recent five years. The research theme was first classified into four following areas such as foodservice, culinary, food and so forth. Then it was subdivided under 65 items on 19 subject matters. Most of the studies are related to foodservice area(595 papers, 56.4%), followed by culinary area (250 papers, 23.7%), food area(105 papers, 10.0%) and other areas (104 papers, 9.9%). For the topics used in foodservice research, most papers investigated the marketing area(170 paper, 28.6%). In culinary and food-related research, papers dealt with confectionary and bread (67 papers, 23.3%), sauce·dressing (38 papers, 13.2%), vegetable·fruit (36 papers, 12.5%), fermented food (35 papers, 12.5). In regard to the analysis of research subjects, institutions, and companies, 140papers(38.2%) conducted hotel-related research, followed by 106 restaurant-related papers (29%) and forty-three papers(11.8%) on foodservice enterprises and franchises. Based on this analysis, most of the theses in ``The Korean Journal of Culinary Research`` have been published in the area of foodservice (56.4%). In comparison, the theses in the area of culinary and food are associated with the name of the journal and only33.7 percent of papers have been released. For these reasons, the name of the journal would rather be renamed ``The Korean Journal of Foodservice and Culinary Research``.

영업양도와 제2차 납세의무와의 관계

강명수(Kang Myeong Su) 조선대학교 법학연구원 2003 法學論叢 Vol.10 No.-

A business transfer refers to the transferring of a business that is a functional property as an organic w hole organized by a certain business purpose to a third person. Therefore. the business is not just an aggregate of products and rights but a managing organization that has great value through business activities. As it is not desirable socially as well as by the business proprietors that a business is dismantled due to a change of the business proprietors. commercial law accepts the business itself as a subject of transfer under the fundamental idea of business maintenance Commercial law does not recognize an effect of inclusive succession due to a business transfer. However. it regulates several matters to prevent unexpected damages to the concerned party that may occur due to a business transfer. Questions have been presented about lawful and realistic bona fides of the regulations of business transfer in the commercial law. A common one concerns the performance liability of a business alienee who takes over the business name under Article 42 of commercial law. That is. if the concerned trade is recognized as a business transfer irregardless of intention or contracts of the concerned parties. alienee who takes over the business shall have an unlimited personal responsibility for a creditor of an alienator. However. it is questionable whether many business people know that the business alienee who takes over the business name should have that responsibility. It is certain that the regulation is designed to protect the creditor of the alienator. but it is no七thoughtthat it is balanced with the situation of the alienator. Artic1e 45 of commercial law regulates the term of responsibility of the business alienator. It says that creditors are exempt from obligation due to a business transfer irregardless of the intention of the creditors. but its theoretical background is not certain. Also. Article 374 of commercial law states that when part of an important business is transferred. a special resolution of a general meeting of stockholders is necessary. but it is not easy to decide the range of the important business In addition. business transfer is not just a matter between alienators and alienee. That is. in a business transfer. the transfer becomes a social matter beyond the security of living. The employment succession of workers can be a reason for strife. It is due to the fact tha t there is no wri tten regulation about service Lransfer in the commercial law or the service law. and there are no clear standards to judge this business transfer. Also. the second liability to taxation of business alienee in the tax law has a different principle of the law from that of the commercial law. When the second taxation liability falls to the business alienee due to a business transfer. judicial order should be maintained and property rights of the second taxation liability should not be violated unreasonably. However. as there is an uncertain concept in the tax law. its judicial security and possibility of fulfillment are not guaranteed The situation is not beneficial to the interest of the third party or to the concerned party. It is not believed that the regulation of the present business transfer meets demands such as guaranty of free business activities. enforced maintenance of business. and guaranty of trade security. This study analysed the second taxation liability under the fundamental national tax law to present an effective direction for protection of the alienee who should be responsible for reimbursement and the second tax payment due to the business transfer.