• Genomics & Informatics
• /
• 제18권2호
• /
• pp.20.1-20.4
• /
• 2020

Accessible negative results are relevant for researchers and clinicians not only to limit their search space but also to prevent the costly re-exploration of research hypotheses. However, most biomedical relation extraction datasets do not seek to distinguish between a false and a negative relation among two biomedical entities. Furthermore, datasets created using distant supervision techniques also have some false negative relations that constitute undocumented/ unknown relations (missing from a knowledge base). We propose to improve the distinction between these concepts, by revising a subset of the relations marked as false on the phenotype-gene relations corpus and give the first steps to automatically distinguish between the false (F), negative (N), and unknown (U) results. Our work resulted in a sample of 127 manually annotated FNU relations and a weighted-F1 of 0.5609 for their automatic distinction. This work was developed during the 6th Biomedical Linked Annotation Hackathon (BLAH6).

• 정보기술과데이타베이스저널
• /
• 제6권2호
• /
• pp.19-28
• /
• 1999

Digital image and video libraries require new algorithms for the automated extraction and indexing of salient image features. Eigen values of an image provide one important cue for the discrimination of image content. In this paper we propose a new approach for automated content extraction that allows efficient database searching using eigen values. The algorithm automatically extracts eigen values from the image matrix represented by the covariance matrix for the image. We demonstrate that the eigen values representing shape information and the skewness of its distribution representing complexity provide good performance in image query response time while providing effective discriminability. We present the eigen value extraction and indexing techniques. We test the proposed algorithm of searching by eigen value and its skewness on a database of 100 images.

• 방송공학회논문지
• /
• 제7권3호
• /
• pp.262-270
• /
• 2002

대부분의 영상색인 기법에서는 영상의 전역 특징값을 이용한다. 그러나 이러한 방법은 영상의 지역적인 변화들을 담아내지 못하기 때문에 만족할 만한 격과를 제공하지 못한다. 본 논문에서는 이러한 문제점을 해결하기 위한 방법으로 영상의 특징점(salient point)과 영상분할을 이용하여 중요영역(important region)을 추출하는 새로운 영역기반 영상검색 시스템을 제안한다. 본 논문에서 제안하는 특징점 추출 기법은 기존의 방법과 비교하여 빠르고 정확한 추출 결과를 보여준다. 선택된 영역에서 추출된 칼라와 질감 정보를 이용하여 검색한 결과는 칼라나 질감 정보의 전력 특징값을 이용한 검색 방법의 결과보다 크게 향상됨을 알 수 있었다.

• 한국식품영양과학회지
• /
• 제27권4호
• /
• pp.687-692
• /
• 1998

To use onion(Allium cepa L.) residue as raw materials of food product, yields of onion-juice according to various extracting methods and optimal extracting condition of quercetin and its related glycosides were carried out. Onion juices gained by the methods of pressing, rotary crushing, freeze pressing and enzyme treatment. The yield by the method of enzyme treatment was higher than others. The yields of juice from fresh onion and heat-treated onion(8$0^{\circ}C$/10min) by pressing were 66% and 83%, respectively. Ethanol extraction of onion was efficient at 75$^{\circ}C$ and acetic acid extraction was proper at 3% concentration for 3 hours at 11$0^{\circ}C$. The onion extract was fractionated in the order of hexane, chloroform, ethylacetate and butanol to test flavonoids. The highest abundant flavonoids were found in ethylacetate and butanol fraction.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 2000년도 하계학술대회 논문집 D
• /
• pp.3034-3036
• /
• 2000

Recently the use of Laser Projector is ascending world-widely to make the best effiencies out of the various events & commercials. To produce fixed image through Laser projector, at least 24 times/sec of iteration is required. In this view, its performance is much influenced by coordinates extraction method for high-speed controlling. This paper basically introduces the way of inputting coordinates by extraction directly from outline rather than indicating size and directions of image's outline coordinates which used by renowned US company 'L' and etc.

• ETRI Journal
• /
• 제33권2호
• /
• pp.259-266
• /
• 2011

We present a new method for an automated markerless system to describe, analyze, and classify human gait motion. The automated system consists of three stages: I) detection and extraction of the moving human body and its contour from image sequences, ii) extraction of gait figures by the joint angles and body points, and iii) analysis of motion parameters and feature extraction for classifying human gait. A sequential set of 2D stick figures is used to represent the human gait motion, and the features based on motion parameters are determined from the sequence of extracted gait figures. Then, a k-nearest neighbor classifier is used to classify the gait patterns. In experiments, this provides an alternative estimate of biomechanical parameters on a large population of subjects, suggesting that the estimate of variance by marker-based techniques appeared generous. This is a very effective and well-defined representation method for analyzing the gait motion. As such, the markerless approach confirms uniqueness of the gait as earlier studies and encourages further development along these lines.

• Journal of Semiconductor Engineering
• /
• 제2권1호
• /
• pp.125-129
• /
• 2021

As the interest of deep learning, techniques to control the color of images in image processing field are evolving together. However, there is no clear standard for color, and it is not easy to find a way to represent only the color itself like the color-palette. In this paper, we propose a novel color palette extraction system by chroma fine-tuning with reinforcement learning. It helps to recognize the color combination to represent an input image. First, we use RGBY images to create feature maps by transferring the backbone network with well-trained model-weight which is verified at super resolution convolutional neural networks. Second, feature maps are trained to 3 fully connected layers for the color-palette generation with a generative adversarial network (GAN). Third, we use the reinforcement learning method which only changes chroma information of the GAN-output by slightly moving each Y component of YCbCr color gamut of pixel values up and down. The proposed method outperforms existing color palette extraction methods as given the accuracy of 0.9140.

• ETRI Journal
• /
• 제42권6호
• /
• pp.815-826
• /
• 2020

We propose a two-dimensional (2D) scattering-center-extraction (SCE) method using sparse recovery based on the compressive-sensing theory, even with data missing from the received radar cross-section (RCS) dataset. First, using the proposed method, we generate a 2D grid via adaptive discretization that has a considerably smaller size than a fully sampled fine grid. Subsequently, the coarse estimation of 2D scattering centers is performed using both the method of iteratively reweighted least square and a general peak-finding algorithm. Finally, the fine estimation of 2D scattering centers is performed using the orthogonal matching pursuit (OMP) procedure from an adaptively sampled Fourier dictionary. The measured RCS data, as well as simulation data using the point-scatterer model, are used to evaluate the 2D SCE accuracy of the proposed method. The results indicate that the proposed method can achieve higher SCE accuracy for an incomplete RCS dataset with missing data than that achieved by the conventional OMP, basis pursuit, smoothed L0, and existing discrete spectral estimation techniques.

• 한국산업보건학회지
• /
• 제9권2호
• /
• pp.135-144
• /
• 1999

The purposes of this study were to survey types of pretreatment methods adopted by industrial hygiene laboratories in Korea for extracting heavy metals in welding fume samples and to compare performances of two pretreatment methods, the acid extraction and the microwave digestion, in extracting heavy metals contained in the real workplace samples from various welding jobs including arc, argon, and carbon dioxide. A total of 25 analytical chemists in the industrial hygiene laboratories participating the quality control program directed by the Korea Industrial Safety Corporation were interviewed by telephone. For the purpose of comparing performance of extracting heavy metals from real workplace samples, a total of 53 welders from 21 workplaces located in Anyang, Uiwang, and Kunpo areas were sampled from the period of March 22, 1999 to April 20, 1999. It was found that the most frequently adopted method for samples from the quality control program was the acid extraction method(40%) followed by the NIOSH 7300 method(36%). The NIOSH method, however, was the dominant method(36%) for samples from workplace followed by the acid extraction method(28%). In this study, two extraction methods, the acid extraction and the microwave digestion, were compared in terms of recovery rate, accuracy, and precision for both manganese and chromium. Both methods produced comparable results for the samples prepared for the quality control program. In contrast, concentrations of two heavy metals determined from real workplace samples pretreated with the microwave digestion method were statis tically significantly higher, manganese(166%) and chromium (200%), than those of utilizing the acid extraction method. These findings were consistent regardless of types of welding techniques used. The results of this study clearly show the importance of verifying the analytical performances of extraction methods for heavy metals not only for the samples from the quality control program but also from the real world samples collected from welding jobs.

• 디지털융복합연구
• /
• 제17권11호
• /
• pp.233-240
• /
• 2019

사이버 공격을 예측하고 대응하기 위해서 수많은 보안 기업 회사에서는 공격기법의 특성, 수법 유형을 빠르게 파악하고, 이에 대한 Security Intelligence Report(SIR)들을 배포한다. 하지만 각 기업에서 배포하는 SIR들은 방대하며, 형식이 맞춰져 있지 않다. 본 논문은 대량의 비정형한 SIR들에서 정보를 추출하는데 소요되는 시간을 줄이고 효율적으로 파악하기 위해 SIR들에 대해 정형화하고 주요 정보를 추출하기 위해 5가지 분석기술이 적용된 프레임워크를 제안한다. SIR들의 데이터는 정답 라벨이 없기 때문에 비지도 학습방식을 통해 키워드 추출, 토픽 모델링, 문서 요약, 유사문서 검색 총 4가지 분석기술을 제안한다. 마지막으로 SIR들에서 위협 정보 추출하기 위해 데이터를 구축하였으며, 개체명 인식 기술에 적용하여 IP, Domain/URL, Hash, Malware에 속하는 단어를 인식하고 그 단어가 어떤 유형에 속하는지 판단하는 분석기술을 포함한 총 5가지 분석기술이 적용된 프레임워크를 제안한다.