• 대한원격탐사학회지
• /
• 제18권6호
• /
• pp.369-376
• /
• 2002

본 연구에서는 공간 영상 자료의 감독 분류에 있어, 분석자에 의하여 선정된 분류 항목별 교사 자료를 분광 특징별로 다수의 군집으로 분리하고, 각각의 군집을 새로운 분류 항목의 교사 자료로서 설정함으로써 분류 성능을 향상시킬 수 있는 기법을 제안하고자 한다 특징 분리를 통하여 생성된 교사 자료는 비교적 작은 값의 밴드별 분산값을 가질 뿐 아니라 정규분포 형태의 자료 분포를 보이게 되어 통계적 감독 분류 기법의 적용에 적합한 교사 자료로서의 성격을 가지게 된다. 제안된 기법은 부산 지역에 대한 Landsat TM 영상 자료를 이용하여 그 적용성이 시험되었으며, 기존의 통계적 분류 기법들에 의한 결과와 그 성능이 정성적으로 비교되었다. 시험 적용 결과, 본 기법은 분석자가 선정한 교사 자료의 분광적인 분포 형태에 관계없이 우수한 분류 성능을 나타내는 것으로 판단되며, 따라서 분류 항목의 설정 및 항목별 교사 자료의 선정에 있어 교사 자료의 분광적 특징에 대한 동일성을 유지하기 위한 노력을 줄여줄 것으로 기대된다.

• 한국음향학회지
• /
• 제40권5호
• /
• pp.439-451
• /
• 2021

본 논문에서는 실제 환경에서의 연속 음성 강화를 위한 딥러닝 기반 온라인 빔포밍 알고리듬과 스펙트럼 감산을 결합한 빔포머를 제안한다. 기존 빔포밍 시스템은 컴퓨터에서 음성과 잡음을 완전히 겹친 방식으로 혼합하여 생성된 사전 분할 오디오 신호를 사용하여 대부분 평가되었다. 하지만 실제 환경에서는 시간 축으로 음성 발화가 띄엄띄엄 발성되기 때문에, 음성이 없는 잡음 신호가 시스템에 입력되면 기존 빔포밍 알고리듬의 성능이 저하된다. 이러한 효과를 경감하기 위하여, 심층 학습 기반 온라인 빔포밍 알고리듬과 스펙트럼 감산을 결합하였다. 잡음 환경에서 온라인 빔포밍 알고리듬을 평가하기 위해 연속 음성 강화 세트를 구성하였다. 평가 세트는 CHiME3 평가 세트에서 추출한 음성 발화와 CHiME3 배경 잡음 및 MUSDB에서 추출한 연속 재생되는 배경음악을 혼합하여 구성되었다. 음성인식기로는 Kaldi 기반 툴킷 및 구글 웹 음성인식기를 사용하였다. 제안한 온라인 빔포밍 알고리듬 과 스펙트럼 감산이 베이스라인 빔포밍 알고리듬에 비해 성능 향상을 보임을 확인하였다.

• 대한수학회보
• /
• 제54권5호
• /
• pp.1669-1675
• /
• 2017

The group of automorphisms of the first Weyl algebra acts on commuting ordinary differential operators with polynomial coefficient. In this paper we prove that for fixed generic spectral curve of genus two the set of orbits is infinite.

• Journal of applied mathematics & informatics
• /
• 제4권1호
• /
• pp.299-308
• /
• 1997

In this paper we show that the limit of a convergent in-vertible sequence in the set of invertible elements Inv(A) in a Banach algebra A under a certain conditions is invertible and we investigate some properties of the spectral radius of banach algebra with unit.

• 음성과학
• /
• 제11권3호
• /
• pp.131-142
• /
• 2004

Speaker identification system basically functions by comparing spectral energy of an individual production model with that of an input signal. This study aimed to develop a new speaker identification system from two parameters from the spectral energy of numeric sounds: difference sum and correlation coefficient. A narrow-band spectrogram yielded more stable spectral energy across time than a wide-band one. In this paper, we collected empirical data from four male speakers and tested the speaker identification system. The subjects produced 18 combinations of three-digit numeric. sounds !en times each. Five productions of each three-digit number were statistically averaged to make a model for each speaker. Then, the remaining five productions were tested on the system. Results showed that when the threshold for the absolute difference sum was set to 1200, all the speakers could not pass the system while everybody could pass if set to 2800. The minimum correlation coefficient to allow all to pass was 0.82 while the coefficient of 0.95 rejected all. Thus, both threshold levels can be adjusted to the need of speaker identification system, which is desirable for further study.

• 대한수학회지
• /
• 제44권4호
• /
• pp.863-872
• /
• 2007

Let D be an integral domain with quotient field K, A a nonempty set of height-one maximal t-ideals of D, F$({\Lambda})={I{\subseteq}D|I$ is an ideal of D such that $I{\subseteq}P$ for all $P{\in}A}$, and $D_F({\Lambda})={x{\in}K|xA{\subseteq}D$ for some $A{\in}F({\Lambda})}$. In this paper, we prove that if each $P{\in}A$ is the radical of a finite type v-ideal (resp., a principal ideal), then $D_{F({\Lambda})}$ is a weakly Krull domain (resp., generalized weakly factorial domain) if and only if the intersection $D_{F({\Lambda})}={\cap}_{P{\in}A}D_P$ has finite character, if and only if $F({\Lambda})$ is a t-splitting set of ideals, if and only if $F({\Lambda})$ is v-finite.

• 대한전자공학회:학술대회논문집
• /
• 대한전자공학회 2003년도 하계종합학술대회 논문집 Ⅳ
• /
• pp.2343-2346
• /
• 2003

The spatial resolution of multispectral images can be improved by merging them with higher resolution image data. A fundamental problem frequently occurred in existing fusion processes, is the distortion of spectral information. This paper presents a spatially adaptive image fusion algorithm which produces visually natural images and retains the quality of local spectral information as well. High frequency information of the high resolution image to be inserted to the resampled multispectral images is controlled by adaptive gains to incorporate the difference of local spectral characteristics between the high and the low resolution images into the fusion. Each gain is estimated to minimize the l$_2$-norm of the error between the original and the estimated pixel values defined in a spatially adaptive window of which the weight are proportional to the spectral correlation measurements of the corresponding regions. This method is applied to a set of co-registered Landsat7 ETM＋ panchromatic and multispectral image data.

• 대한기계학회논문집A
• /
• 제27권2호
• /
• pp.294-301
• /
• 2003

It is of often important to accurately predict the flow-induced vibration or dynamic instability of a pipeline conveying internal high speed flow in advance, which requires a very accurate solution method. In this study, first the dynamic equations for the axial and transverse vibrations of a pipeline are reduced from a set of pipe-dynamic equations derived in the previous study and then the spectral element model is formulated. The accuracy of the spectral element method (SEM) is then verified by comparing its results with the results obtained by finite element method (FEM). It is shown that the present spectral element model provides very accurate solutions by using an extremely small number of degrees-of-freedom when compared with FEM. The dynamics of a sample pipeline is investigated with varying the axial tension and the speed of internal flow.

• 한국축산시설환경학회지
• /
• 제20권3호
• /
• pp.117-124
• /
• 2014

The objectives were to assess the use of near-infrared reflectance spectroscopy (NIRS) as a tool for estimating nutrient compositions of corn kernel, and to apply an NIRS-based indium gallium arsenide array detector to the system for collecting spectra and analyzing calibration equations using equipments designed for field application. Partial Least Squares Regression (PLSR) was employed to develop calibration equations based on representative data sets. The kennard-stone algorithm was applied to induce a calibration set and a validation set. As a result, the method for structuring a calibration set significantly affected prediction accuracy. The prediction of chemical composition of corn kernel resulted in the following (kennard-stone algorithm: relative) moisture ($R^2=0.82$, RMSEP=0.183), crude protein ($R^2=0.80$, RMSEP=0.142), crude fat ($R^2=0.84$, RMSEP=0.098), crude fiber ($R^2=0.74$, RMSEP=0.098), and crude ash ($R^2=0.81$, RMSEP=0.048). Result of this experiment showed the potential of NIRS to predict the chemical composition of corn kernel.

• Journal of Radiation Protection and Research
• /
• 제42권2호
• /
• pp.91-97
• /
• 2017

Background: A gamma energy identifying algorithm using spectral decomposition combined with smoothing method was suggested to confirm the existence of the artificial radio isotopes. The algorithm is composed by original pattern recognition method and smoothing method to enhance the performance to identify gamma energy of radiation sensors that have low energy resolution. Materials and Methods: The gamma energy identifying algorithm for the compact radiation sensor is a three-step of refinement process. Firstly, the magnitude set is calculated by the original spectral decomposition. Secondly, the magnitude of modeling error in the magnitude set is reduced by the smoothing method. Thirdly, the expected gamma energy is finally decided based on the enhanced magnitude set as a result of the spectral decomposition with the smoothing method. The algorithm was optimized for the designed radiation sensor composed of a CsI (Tl) scintillator and a silicon pin diode. Results and Discussion: The two performance parameters used to estimate the algorithm are the accuracy of expected gamma energy and the number of repeated calculations. The original gamma energy was accurately identified with the single energy of gamma radiation by adapting this modeling error reduction method. Also the average error decreased by half with the multi energies of gamma radiation in comparison to the original spectral decomposition. In addition, the number of repeated calculations also decreased by half even in low fluence conditions under $10^4$ ($/0.09cm^2$ of the scintillator surface). Conclusion: Through the development of this algorithm, we have confirmed the possibility of developing a product that can identify artificial radionuclides nearby using inexpensive radiation sensors that are easy to use by the public. Therefore, it can contribute to reduce the anxiety of the public exposure by determining the presence of artificial radionuclides in the vicinity.