• 제어로봇시스템학회논문지
• /
• 제19권12호
• /
• pp.1081-1088
• /
• 2013

Hyperspectral cameras acquire reflectance values at many different wavelength bands. Dimensions tend to increase because spectral information is stored in each pixel. Several attempts have been made to reduce dimensional problems such as the feature selection using Adaboost and dimension reduction using the Simulated Annealing technique. We propose a novel material detection method that consists of four steps: feature band selection, feature extraction, SVM (Support Vector Machine) learning, and target and specific region detection. It is a combination of the band ratio method and Simulated Annealing algorithm based on detection rate. The experimental results validate the effectiveness of the proposed feature selection and band ratio method.

• 한국멀티미디어학회논문지
• /
• 제25권6호
• /
• pp.819-832
• /
• 2022

Cross-domain recommender system transfers knowledge across different domains to improve the recommendation performance in a target domain that has a relatively sparse model. However, they suffer from the "negative transfer" in which transferred knowledge operates as noise. This paper proposes a novel Multiple Fusion-based Deep Cross-Domain Recommendation named MFDCR. We exploit Doc2Vec, one of the famous word embedding techniques, to fuse data user-wise and transfer knowledge across multi-domains. It alleviates the "negative transfer" problem. Additionally, we introduce a simple multi-layer perception to learn the user-item interactions and predict the possibility of preferring items by users. Extensive experiments with three domain datasets from one of the most famous services Amazon demonstrate that MFDCR outperforms recent single and cross-domain recommendation algorithms. Furthermore, experimental results show that MFDCR can address the problem of "negative transfer" and improve recommendation performance for multiple domains simultaneously. In addition, we show that our approach is efficient in extending toward more domains.

• 한국진공학회:학술대회논문집
• /
• 한국진공학회 2011년도 제40회 동계학술대회 초록집
• /
• pp.421-422
• /
• 2011

The first part is about development of a liquid target for a neutron source, which is designed to overcome many of the limitations of traditional beam-target neutron generators by utilizing a liquid target neutron source. One of the most critical aspects of the beam-target neutron generator is the target integrity under the beam exposure. A liquid target can be a good solution to overcome damage to the target such as target erosion and depletion of hydrogen isotopes in the active layer, especially for the one operating at high neutron fluxes with no need for water cooling. There is no inherent target lifetime for the liquid target neutron generator when used with continuous refreshment of the target surface exposed to the energetic beam. In this work, liquid target containing hydrogen has been developed and tested in vacuum environment. Potentially, liquid targets could allow a point neutron source whose spatial extension is on the order of 1 to $10{\mu}m$. And the second is about the vacuum ultraviolet (VUV) spectrometer which is designed as a five-channel spectral system for ITER main plasma measurement. To develop and verify the design, a two-channel prototype system was fabricated with No. 3 (14.4 nm~31.8 nm) and No. 4 (29.0 nm~60.0 nm) among the five channels. For test of the prototype system, a hollow cathode lamp is used as a light source. The system is composed of a collimating mirror to collect the light from source to slit, and two holographic diffraction gratings with toroidal geometry to diffract and also to collimate the light from the common slit to detectors. The two gratings are positioned at different optical distances and heights as designed. To study the appropriate detector for ITER VUV system, two different electronic detectors of the back-illuminated charge coupled device and the micro-channel plate electron multiplier were installed and the performance has been investigated and compared in the same experimental conditions. The overall system performance was verified by measuring the spectrums.

• 제어로봇시스템학회:학술대회논문집
• /
• 제어로봇시스템학회 1997년도 한국자동제어학술회의논문집; 한국전력공사 서울연수원; 17-18 Oct. 1997
• /
• pp.524-527
• /
• 1997

Conventional target tracking algorithms based on the linear estimation techniques perform quite efficiently when the target motion does not involve maneuvers. Target maneuvers involving short term accelerations, however, cause a bias in the measurement sequence. Accurate compensation for the bias requires processing more samples of which adds to the computational complexity. The primary motivation for employing a neural network for this task comes from the efficiency with which more features can be as inputs for bias compensation. A system architecture that efficiently integrates the fusion capabilities of a trained multilayer neural net with the tracking performance of a Kalman filter is described. The parallel processing capability of a properly trained neural network can permit fast processing of features to yield correct acceleration estimates and hence can take the burden off the primary Kalman filter which still provides the target position and velocity estimates.

• BMB Reports
• /
• 제43권5호
• /
• pp.319-324
• /
• 2010

Experimental bioinformatics data obtained from an E. coli cell-based eukaryotic protein purification experiment were analyzed in order to identify any bottleneck as well as the factors affecting the target purification. All targets were expressed as His-tagged maltose-binding protein (MBP) fusion constructs and were initially purified by immobilized metal affinity chromatography (IMAC). The targets were subsequently separated from the His-tagged MBP through TEV protease cleavage followed by a second IMAC isolation. Of the 743 total purification trials, 342 yielded more than 3 mg of target proteins for structural studies. The major reason for failure of target purification was poor TEV proteolysis. The overall success rate for target purification decreased linearly as cysteine content or isoelectric point (pI) of the target increased. This pattern of pI versus overall success rate strongly suggests that pI should be incorporated into target scoring criteria with a threshold value.

• Journal of Advanced Marine Engineering and Technology
• /
• 제29권5호
• /
• pp.560-570
• /
• 2005

In this paper, we discuss the tracking system for a wheelchair which can follow the path of a human target such as a nurse in hospital. The problem of human tracking is that it requires recognition of feature as well as the tracking of human positions. For this purpose the use of a high cost visual sensor such as laser finder or streo camera makes the tracking a high cost additional expense. This paper proposes the tracking system uses a low cost infrared range-finder and CCD camera, The Infrared range-finder and CCD camera can create a target candidate through each target recognition algorithm. and this information is fused in order to reduce the uncertainties of a target decision and correct the positional error of the human. The effectiveness of the proposed system is verified through experiments.

• 한국재료학회지
• /
• 제26권4호
• /
• pp.222-227
• /
• 2016

Tin oxides have been studied for various applications such as gas detecting materials, transparent electrodes, transparent devices, and solar cells. p-type SnO is a promising transparent oxide semiconductor because of its high optical transparency and excellent electrical properties. In this study, we fabricated p-type SnO thin film using rf magnetron sputtering with an SnO/Sn composite target; we examined the effects of various oxygen flow rates on the SnO thin films. We fundamentally investigated the structural, optical, and electrical properties of the p-type SnO thin films utilizing X-ray diffraction (XRD), X-ray photoelectron spectroscopy (XPS), UV/Vis spectrometry, and Hall Effect measurement. A p-type SnO thin film of $P_{O2}=3%$ was obtained with > 80% transmittance, carrier concentration of $1.12{\times}10^{18}cm^{-3}$, and mobility of $1.18cm^2V^{-1}s^{-1}$. With increasing of the oxygen partial pressure, electrical conductivity transition from p-type to n-type was observed in the SnO crystal structure.

• 한국측량학회지
• /
• 제34권4호
• /
• pp.349-356
• /
• 2016

This study investigates two camera self-calibration approaches, on-site self-calibration and laboratorial self-calibration, both of which are based on self-calibration theory and implemented by using a commercial photogrammetric solution, Agisoft PhotoScan. On-site self-calibration implements camera self-calibration and aerial triangulation by using the same aerial photos. Laboratorial self-calibration implements camera self-calibration by using photos captured onto a patterned target displayed on a digital panel, then conducts aerial triangulation by using the aerial photos. Aerial photos are captured by an unmanned aerial vehicle, and target photos are captured onto a 27in LCD monitor and a 47in LCD TV in two experiments. Calibration parameters are estimated by the two approaches and errors of aerial triangulation are analyzed. Results reveal that on-site self-calibration excels laboratorial self-calibration in terms of vertical accuracy. By contrast, laboratorial self-calibration obtains better horizontal accuracy if photos are captured at a greater distance from the target by using a larger display panel.

• 한국전자파학회논문지
• /
• 제28권3호
• /
• pp.217-224
• /
• 2017

본 논문에서는 항공기 식별을 위해 서로 보완적인 특성을 갖는 HRRP 구분기와 JEM 구분기를 융합하여 식별하는 방법을 제안한다. 다양한 상황에서 단일 구분기보다 향상된 식별성능을 얻기 위하여 식별결과에 대한 구분기의 신뢰도를 가중치로 융합하는 방법을 제안한다. 신뢰도는 구분기의 식별성능으로부터 추정된 사후확률로 정의되며 식별결과에 대한 확신도 및 관측각도에 따라 변하는 특성을 가진다. 시뮬레이션 데이터를 사용한 식별실험을 통해 제안한 융합 방법이 단일 구분기를 효과적으로 융합하여 향상된 식별성능을 얻을 수 있음을 확인하였다.

• Asian Pacific Journal of Cancer Prevention
• /
• 제15권5호
• /
• pp.2309-2312
• /
• 2014

Esophageal squamous cell carcinoma (ESCC) is the most common histologic subtype of esophageal cancer and is characterized by a poor prognosis. Determining gene changes in ESCCs should improve understanding of putative risk factors and provide potential targets for therapy. We sequenced about 55 million pair-end reads from a pair of adjacent normal and ESCC samples to identify the gene expression level and gene fusion. Sanger sequencing was used to verify the result. About 17 thousand genes were expressed in the tissues, of which approximately 2400 demonstrated significant differences between tumor and adjacent non tumor tissue. GO and KEGG pathway analysis revealed that many of these genes were associated with cellular adherence and movement, simulation responses and immune responses. Notably we identified and validated one fusion gene, HLA-E and HLA-B, located 1 MB apart. We also identified thousands of remarkably expressed transcripts. In conclusion, a novel fusion gene HLA-E and HLA-B was identified in ESCC via whole transcriptome sequencing, which would be a biomarker for ESCC diagnosis and target for therapy, shedding new light for better understanding of ESCC tumorigenesis.