http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
GSC 기반 빔포밍을 위한 주파수 밴드별 전력비 분포의 혼합 가우시안 모델을 이용한 목표 음성신호의 검출
장형욱,김영일,정상배,Chang, Hyungwook,Kim, Youngil,Jeong, Sangbae 한국정보통신학회 2015 한국정보통신학회논문지 Vol.19 No.1
다양한 종류의 잡음에 의해서 발생하는 음성인식 성능 저하를 보상하기 위해서는 잡음제거가 필수적이다. 마이크로폰 배열을 이용하는 많은 잡음제거 기술 중에서, GSC는 비정상성 잡음을 제거하기 위해서 널리 적용되어 왔다. GSC의 성능은 AMC에 의해서 직접적인 영향을 받는다. 즉, 정확한 목표 음성 신호의 검출은 순수 잡음구간에서의 충분한 잡음제거 및 목표 음성구간에서의 적은 왜곡을 보장하기 위해서 필수적이다. 따라서, 본 논문에서는 고정 빔포밍의 출력과 차단 매트릭스의 출력간의 전력비가 주파수 밴드 단위로 계산되는 향상된 AMC 설계법을 제안한다. 그 후, 밴드별 전력비는 가우시안 혼합에 의해서 각 클래스가 확률적으로 모델링 된다. 실험결과, 제안한 알고리즘이 ROC 및 출력 SNR 관점에서 더 높은 성능을 보였다. Noise reduction is necessary to compensate for the degradation of recognition performance by various types of noises. Among many noise reduction techniques using microphone array, generalized sidelobe canceller (GSC) has been widely applied to reduce nonstationary noises. The performance of GSC is directly affected by its adaptation mode controller (AMC). That is, accurate target speech detection is essential to guarantee the sufficient noise reduction in pure noise intervals and the less distortion in target speech intervals. Thus, this paper proposes an improved AMC design technique in which the power ratio of the output of fixed beamforming to that of blocking matrix is calculated frequency bandwise and probabilistically modeled by mixture Gaussians for each class. Experimental results show that the proposed algorithm outperforms conventional AMCs in receiver operating curves (ROC) and output SNRs.
10 Topical Lidocaine과 EMLA 크림을 이용한 고막 국소마취 효과 비교
장형욱 대한이비인후과학회 2003 대한이비인후과학회지 두경부외과학 Vol.46 No.3
Background and Objectives:Minor otologic surgeries, such as tympanic injection, myringotomy and ventilating tube insertion are a common otologic procedure. Effective, safe and comfortable anesthetic method is very useful alternative to lidocaine infiltration anesthesia. This study was designed to assess the efficacy of topical lidocaine as compared with eutectic mixture of local anesthetics (EMLA) for pain relief during minor otologic surgery. Materials and Methods:One hundred eight patients who were between the age of 4 and 71 years were allocated randomly into 4 groups. Group I (n=30) received 1 spary of topical lidocaine;group II (n=34), 2 sprays of topical lidocaine;group III (n=23), 1 cc of EMLA cream;and group IV (n=21), 2 cc of EMLA cream to the ear canal. Groups were randomized to perform one of otologic minor surgery at anesthetic application time of either 10 or 20 minutes respectively. Patients rated pain and satisfaction during surgery on the comfort scale score. Results:All surgeries were successfully without additional anesthesia. At 10 minute application, group I and II showed significantly lower comfort scores of pain compared with group III (p<0.05). Group II showed significantly lower scores than group III at 20 minute application (p<0.05). In lidocaine applied group I and II comfort scores of pain were significantly lower at 20 minute application than 10 minute application (p<0.05). There was one complication of severe vertigo after one hour of procedure in group II. Conclusion:This study demonstrates that 10 minute application of topical lidocaine is more effective for relieving pain associated with minor otologic surgery in office than 10 minute application of EMLA cream and 20 minute application of lidocaine is more effective than 10 minute application. (Korean J Otolaryngol 2003;46:207-10)
장형욱,이재진,변경진,엄낙웅,정상배,Jang, HyeongUk,Lee, Jae-Jin,Byun, Kyungjun,Eum, Nakwoong,Jeong, Sangbae 한국스마트미디어학회 2014 스마트미디어저널 Vol.3 No.4
버츄얼 플랫폼은 SoC를 구성하는 프로세서 코어 및 주변장치들을 소프트웨어로 모델링한 것으로, 현재 국내외 대기업에서는 버츄얼 플랫폼을 활용한 Top-Down 설계 플로우를 기반으로 최적 SW+SoC 융합시스템 구조 설계 및 IP 재활용을 통해 개발한 다양한 플랫폼을 제품 개발에 활용하고 있다. 본 논문에서는 오픈 IP인 OpenRISC 프로세서 코어 기반의 버츄얼 플랫폼을 제안한다. 제안된 버츄얼 플랫폼은 타겟 코드를 호스트 코드로 변환하여 수행하는 코드 변환 기법을 사용하여 약 20 MIPS 급의 고속 에뮬레이션을 지원한다. A virtual platform models a processor core and the peripheral devices constituting the SoC in software. Major companies utilize a variety of platforms for product development with optimal SW+SoC integrated system architecture design and IP reuse based Top-Down design flow using a virtual platform. In this paper, we propose a virtual platform based on OpenRISC, an open source RISC based core. The proposed virtual platform supports high speed emulation of approximately 20 MIPS using DBT (Dynamic Binary Translation).
인체 유양동 점막에서 정량적 Real Time RT-PCR을 이용한 점액 유전자의 발현
장형욱,신동원,신승헌,손수준,전창호,이상흔 대한이비인후과학회 2003 대한이비인후과학회지 두경부외과학 Vol.46 No.8
Background and Objectives:Mucin is involved in the pathogenesis of otitis media, but little is known about its expresion in the mastoid mucosa with chronic otitis media. The purpose of this study is to identify MUC1, 2, 4, 5AC, 5B, 8 gene expression in the mastoid mucosa with chronic otitis media, and to compare the results with normal mastoid mucosa. Materials and Method:Normal or inflamed mastoid mucosal specimen were taken at the time of surgery, who underwent cochlear implant surgery and middle ear surgery for chronic otitis media. The profiles of mucin gene expression were examined by Real time reverse trans-cription polymerase chain reaction. Results:This study revealed that RT-PCR of cDNAs from mastoid mucosa specimen of chronic otitis media showed the expression of MUC1, MUC4, MUC5AC, MUC5B, and MUC8 but MUC2 was not detected while normal mastoid mucosa showed the same pattern of mucin gene expression. In mastoid mucosa with chronic otitis media, ± 1.1, MUC4 3.2± 2.5, MUC5AC 2.5± 1.0, MUC5B 4.1± 2.2, and MUC8 2.7± 2.1 times compared with that of normal mucosa and MUC5B is statisticaly significant (p<0.05). The MUC1, 4, 5AC, 5B, 8 mRNA expressions were correlated to infiltration of inflamatory cells in the submucosa of mastoid mucosa with chronic otitis media. Conclusion:These results sugest that MUC1, MUC4, MUC5AC, MUC5B, and MUC8 are expressed in the normal and inflamed mastoid mucosa and up-regulated by chronic inflammation. (Korean J Otolaryngol 2003 ;46 :634-40)
MUSIC 및 반향 성분 제거 기법을 이용한 음성신호의 입사각 추정
장형욱,정상배,김영일,Chang, Hyungwook,Jeong, Sangbae,Kim, Youngil 한국정보통신학회 2014 한국정보통신학회논문지 Vol.18 No.6
본 논문에서는 다중 신호 분류 기반의 음성신호의 입사각 추정 향상 방법을 제안한다. 기본적으로 제안한 방식은 복소 대역통과 필터를 이용하여 신호 분석을 위한 협대역 신호를 생성한다. 또한, 공간 스펙트럼에서의 반향 성분 제거 및 2차 함수 기반의 응답 근사화를 사용하여 추정 각도의 정확도를 향상시켰다. 실험결과 제안한 방법은 일반화된 상호상관도 방식의 입사각 추정 알고리즘보다 검출 오차 및 검출 성공률 측면에서 더 좋은 성능을 보였다. In this paper, we propose a method to improve the performance of the direction-of-arrival (DOA) estimation of a speech source using a multiple signal classification (MUSIC)-based algorithm. Basically, the proposed algorithm utilizes a complex coefficient band pass filter to generate the narrow band signals for signal analysis. Also, reverberation component reduction and quadratic function-based response approximation in MUSIC spatial spectrum are utilized to improve the accuracy of DOA estimation. Experimental results show that the proposed method outperforms the well-known generalized cross-correlation (GCC)-based DOA estimation algorithm in the aspect of the estimation error and success rate, respectively.Abstract should be placed here. These instructions give you guidelines for preparing papers for JICCE.