• 한국통신학회논문지
• /
• 제22권3호
• /
• pp.539-546
• /
• 1997

In this paper, we proposed a new mehtod to analyze the performance degradation by timing jitters in the APD (avalanche photodiode) receivers of intensity modulation/direct detection digital optical communication systems where raised cosine pulse-shaping filters are used to reduce the effect of noise while minimizing intersymbol interferences. The proposed analytical method is an extension of an analytical method we have already developed for pin diode receivers, and incorporates the effects of APD's multiplication factor and resulting shot noise. Using the proposed analytical method, we derive an approximated power penalty due to timing jitters based on an assumption of Gaussian distribution for timing jitters, and compare with that of the conventional analytical method. The results obtained from the proposed analytical method show that conventional analytical methods underestimate the influence of timing jitters on the reciver performance. The results also show that APD's multiplication factor which optimizes receiver sensitivity is smaller than that obtained by the conventional analytical method.

• 한국소음진동공학회:학술대회논문집
• /
• 한국소음진동공학회 2001년도 추계학술대회논문집 II
• /
• pp.1040-1043
• /
• 2001

This paper describes the principle behind Crank Angle Analysis, as implemented by Bruel ＆ Kjaer in the Non-Stationary Spatial Transformation of Sound Fields (NS-STSF) system. The NS-STSF system combines a Time Domain Holography measurement on for example an engine with two simultaneously recorded Tacho signals. The Tacho signals provide the crankshaft angle and the RPM at the instant of each instantaneous output (snap-shot) from Time Domain Holography. As a result, the system allows precise analysis of the temporal and spatial relation between the acoustical emission (or the vibration pattern) and the mechanical events during an engine cycle. Some results from a measurement on a DaimlerChrysler engine are presented.

• 한국지구물리탐사학회:학술대회논문집
• /
• 한국지구물리탐사학회 2005년도 공동학술대회 논문집
• /
• pp.223-226
• /
• 2005

동일한 파원에서 발생된 P, SH, SV 파를 3성분 지오폰으로 동시에 기록하는 방법을 연구하기 위해서, 94 m의 측선을 따라 2 m 간격으로 설치한 수신점에서 기록된 24채널 자료를 512 ms동안 기록하였다. 원시자료의 수직성분에는 P파가 초동으로 뚜렷하게 나타나며, 측선방향과 수직한 수평성분은 P파가 약화된 반면 S파의 신호/잡음비가 향상된 모습을 보인다. 토모그래피 역산을 통해 계산된 P파 및 S파의 속도 토모그램을 이용하여 동포아송비의 2차원적 분포를 산출한 결과, 동포아송비는 $0.2{\~}0.3$의 범위로 분포하며, 이는 한번의 타격으로 P파 및 S파의 속도분포와 더불어 및 동포아송비의 2차원적 분포까지 획득할 수 있다는데 큰 의의가 있다고 판단된다.

• 한국광학회지
• /
• 제3권3호
• /
• pp.161-166
• /
• 1992

회전하는 거친금속표면에 s-편극된 레이저광을 $-30^{\circ}$로 입시시켜 생긴 후방산란파중 입사광과 동일한 편극을 가지는 후방산란파에 대하여 세기의 시간상관함수 $g_{s}^{(2)}(\tau)$를 측정하였다. 이 과정에서 광전신호로부터 $g_{s}^{(2)}(0)$의 측정을 할 때 수반되는 shot noise의 영향을 제하여 줌으로써 정확한 $g_{s}^{(2)}(0)$ 값을 얻었다. 각 산란각$\theta_{s}$에서$g_{s}^{(2)}(\tau)$는 {1+exp($-\tau^2/\tau_0^2$)}함수와 거의 일치하였는데, 이는 등속으로 회전하는 간유리를 투과하여 형성된 레이저 스펙클의 경우와 같은 결과이다. 또한, 산란각에 따른 $\tau_0$의 분포에서는 $\theta_s=34^{\circ}$에서 예리하게 최대값을 가지는 것이 관찰되었다. 이 연구에서, 다중산란이 10%를 넘는 금속판산란과 다중산란이 무시될 수 있는 간유리판산란이 같은 함수꼴의 세기의 시간상관함수를 갖는다는 사실이 확인되었다.

• 천문학회보
• /
• 제44권1호
• /
• pp.54.2-54.2
• /
• 2019

Removing noise which occurs inevitably when taking image data has been a big concern. There is a way to raise signal-to-noise ratio and it is regarded as the only way, image stacking. Image stacking is averaging or just adding all pixel values of multiple pictures taken of a specific area. Its performance and reliability are unquestioned, but its weaknesses are also evident. Object with fast proper motion can be vanished, and most of all, it takes too long time. So if we can handle single shot image well and achieve similar performance, we can overcome those weaknesses. Recent developments in deep learning have enabled things that were not possible with former algorithm-based programming. One of the things is generating data with more information from data with less information. As a part of that, we reproduced stacked image from single shot image using a kind of deep learning, conditional generative adversarial network (cGAN). r-band camcol2 south data were used from SDSS Stripe 82 data. From all fields, image data which is stacked with only 22 individual images and, as a pair of stacked image, single pass data which were included in all stacked image were used. All used fields are cut in $128{\times}128$ pixel size, so total number of image is 17930. 14234 pairs of all images were used for training cGAN and 3696 pairs were used for verify the result. As a result, RMS error of pixel values between generated data from the best condition and target data were $7.67{\times}10^{-4}$ compared to original input data, $1.24{\times}10^{-3}$. We also applied to a few test galaxy images and generated images were similar to stacked images qualitatively compared to other de-noising methods. In addition, with photometry, The number count of stacked-cGAN matched sources is larger than that of single pass-stacked one, especially for fainter objects. Also, magnitude completeness became better in fainter objects. With this work, it is possible to observe reliably 1 magnitude fainter object.

• 방송공학회논문지
• /
• 제26권7호
• /
• pp.844-854
• /
• 2021

Semi-supervised learning (SSL) and few-shot learning (FSL) have shown impressive performance even then the volume of labeled data is very limited. However, SSL and FSL can encounter a significant performance degradation if the diversity gap between the labeled and unlabeled data is high. To reduce this diversity gap, we propose a novel scheme that relies on an autoencoder for generating pseudo examples. Specifically, the autoencoder is trained on a specific class using the available labeled data and the decoder of the trained autoencoder is then used to generate N samples of that specific class based on N random noise, sampled from a standard normal distribution. The above process is repeated for all the classes. Consequently, the generated data reduces the diversity gap and enhances the model performance. Extensive experiments on MNIST and FashionMNIST datasets for SSL and FSL verify the effectiveness of the proposed approach in terms of classification accuracy and robustness against adversarial attacks.

• Corrosion Science and Technology
• /
• 제12권2호
• /
• pp.93-101
• /
• 2013

The object of this work is to establish an electrochemical noise(EN) measurement technique combined with a direct current potential drop(DCPD) method for monitoring of localized corrosion cracking of nickel-based alloy, and to analyze its mechanism. The electrochemical current and potential noises were measured under various conditions of applied stress to a compact tension specimen in a simulated primary water chemistry of a pressurized water reactor. The amplitude and frequency of the EN signals were evaluated in both time and frequency domains based on a shot noise theory, and then quantitatively analyzed using statistical Weibull distribution function. From the spectral analysis, the effect of the current application in DCPD was found to be effectively excluded from the EN signals generated from the localized corrosion cracking. With the aid of a microstructural analysis, the relationship between EN signals and the localized corrosion cracking mechanism was investigated by comparing the shape parameter of Weibull distribution of a mean time-to-failure.

• 스마트미디어저널
• /
• 제10권1호
• /
• pp.79-87
• /
• 2021

In recent years research on automatic font generation with machine learning mainly focus on using transformation-based methods, in comparison, generative model-based methods of font generation have received less attention. Transformation-based methods learn a mapping of the transformations from an existing input to a target. This makes them ambiguous because in some cases a single input reference may correspond to multiple possible outputs. In this work, we focus on font generation using the generative model-based methods which learn the buildup of the characters from noise-to-image. We propose a novel way to train a conditional generative deep neural model so that we can achieve font style control on the generated font images. Our research demonstrates how to generate new font images conditioned on both character class labels and character style labels when using the generative model-based methods. We achieve this by introducing a modified generator network which is given inputs noise, character class, and style, which help us to calculate losses separately for the character class labels and character style labels. We show that adding the character style vector on top of the character class vector separately gives the model rich information about the font and enables us to explicitly specify not only the character class but also the character style that we want the model to generate.

• 천문학회보
• /
• 제46권1호
• /
• pp.37.3-38
• /
• 2021

Radiation pressure noise of photon and photon shot noise are quantum noise limitation in interferometric gravita-tional wave detectors. Since relationship between the two noises is position and momentum of the Heisenberg uncertainty principle, quantum non-demolition (QND) technique is required to reduce the two noises at the same time. Frequency dependent squeezing using a filter cavity is one of realistic solutions for QND measurement and experimental results show that its cutting-edge performance is sufficient to apply to the current gravitational wave detectors. A 300m filter cavity is under construction at adv-LIGO. KAGRA (gravitational wave detector in Japan) has also started international collaboration to build a filter cavity. Recently we joined the filter cavity project for KAGRA. Current status of squeezing and filter cavity research at KASI and details of the KAGRA filter cavity project will be presented.

• 반도체디스플레이기술학회지
• /
• 제22권4호
• /
• pp.93-98
• /
• 2023

As semiconductor devices become highly integrated and process difficulty increases, the need for highly sensitive sensors that can detect micro leaks is increasing. However, the noise contained in the CCD sensor itself acts as an obstacle to detecting fine leaks. In this study, integration time was changed for each condition, the sensor was cooled to 0℃, and the dark voltage level was measured to confirm through experiment the characteristics of the temporal noise included in the CCD sensor, a component of OES (Optical Emission Spectroscopy). When integration time was reduced from 30msec to 10msec, the dark voltage level decreased by about 20.5 % from an average of 151.5mV to 120.5mV. In the case of cooling device, Peltier elements were selected because of their simple structure and small size. During temperature cooling, the target temperature was controlled to within ±0.5℃ through PID control. When cooled from 20℃ to 0℃ using this cooling device, it was confirmed that the dark voltage level decreased by about 7% from an average of 147.0mV to 137.0mV.