• 한국정보과학회 언어공학연구회:학술대회논문집(한글 및 한국어 정보처리)
• /
• 한국정보과학회언어공학연구회 2021년도 제33회 한글 및 한국어 정보처리 학술대회
• /
• pp.540-543
• /
• 2021

본 논문에서는 Multi-Head Attention 대신 Spatial Gating Unit을 사용하는 GMLP[1]에 작은 Attention 신경망을 추가한 모델을 구성하여 뉴스와 위키피디아 데이터로 사전학습을 실시하고 한국어 다운스트림 테스크(감성분석, 개체명 인식)에 적용해 본다. 그 결과, 감성분석에서 Multilingual BERT보다 0.27%높은 Accuracy인 87.70%를 보였으며, 개체명 인식에서는 1.6%높은 85.82%의 F1 Score를 나타내었다. 따라서 GMLP가 기존 Transformer Encoder의 Multi-head Attention[2]없이 SGU와 작은 Attention만으로도 BERT[3]와 견줄만한 성능을 보일 수 있음을 확인할 수 있었다. 또한 BERT와 추론 속도를 비교 실험했을 때 배치사이즈가 20보다 작을 때 BERT보다 1에서 6배 정도 빠르다는 것을 확인할 수 있었다.

• IEIE Transactions on Smart Processing and Computing
• /
• 제4권2호
• /
• pp.115-125
• /
• 2015

This paper presents a Cuckoo search algorithm to secure adversaries misdirecting multi-hop routing in Mobile ad hoc networks (MANETs) using a robust Trust Predicated Routing Framework with an optimized cluster head selection. The clustering technique designed in this framework leads to efficient routing in MANETs. The heavy work load in the node causes an energy drop in cluster head, which leads to re-clustering of the group, and another cluster head is selected to avoid packet loss during data transmission. The problem in the re-clustering process is that the overall efficiency of the routing process is reduced and the processing time is increased. A Cuckoo search based optimization algorithm is proposed to solve the problem of re-clustering by selecting the secondary cluster head within the initially formed cluster group and eliminating the reclustering process. The proposed framework enables a node to select a reliable and secure route for MANET and the performance can be evaluated by comparing the simulated results with the AODV routing protocol, which shows that the performance of the proposed routing protocol are improved significantly.

• 임상이비인후과
• /
• 제29권2호
• /
• pp.240-244
• /
• 2018

Granulomatosis with polyangiitis (GPA) and Immunoglobulin (Ig) $G_4$-related disease ($IgG_4$-RD) are rare diseases and early diagnosis and proper management are imperative to prevent multi-organ damage. The authors present a case of a 60 years old woman who had facial paralysis and hearing loss. Lt intact canal wall tympanomastoidectomy, Lt facial nerve decompression and ossiculoplasty with partial ossicular replacement prosthesis (PORP) was done. During operation, middle ear tissue was biopsied and GPA with $IgG_4$-RD was diagnosed. After methyl prednisolone (MPD) pulse therapy and azathioprine therapy, the severity of paralysis was improved. We present this case because common otologic symptoms like facial palsy and hearing loss could be initial symptoms of rare systemic disease.

• Journal of Animal Science and Technology
• /
• 제65권3호
• /
• pp.519-534
• /
• 2023

Rumination in cattle is closely related to their health, which makes the automatic monitoring of rumination an important part of smart pasture operations. However, manual monitoring of cattle rumination is laborious and wearable sensors are often harmful to animals. Thus, we propose a computer vision-based method to automatically identify multi-object cattle rumination, and to calculate the rumination time and number of chews for each cow. The heads of the cattle in the video were initially tracked with a multi-object tracking algorithm, which combined the You Only Look Once (YOLO) algorithm with the kernelized correlation filter (KCF). Images of the head of each cow were saved at a fixed size, and numbered. Then, a rumination recognition algorithm was constructed with parameters obtained using the frame difference method, and rumination time and number of chews were calculated. The rumination recognition algorithm was used to analyze the head image of each cow to automatically detect multi-object cattle rumination. To verify the feasibility of this method, the algorithm was tested on multi-object cattle rumination videos, and the results were compared with the results produced by human observation. The experimental results showed that the average error in rumination time was 5.902% and the average error in the number of chews was 8.126%. The rumination identification and calculation of rumination information only need to be performed by computers automatically with no manual intervention. It could provide a new contactless rumination identification method for multi-cattle, which provided technical support for smart pasture.

• 천문학회지
• /
• 제21권2호
• /
• pp.117-132
• /
• 1988

Multi-frequency polarization observations of Abell 695 are reported here. The brightest radio source 0838+325, which was previously classified as a wide angled head-tail radio source, is, in the present observations, resolved into two separable sources, a head-tail source 0838+325 AB and a diffuse one 0838+325C. The radio-tail shows high degree of linear polarization(${\sim}50$-25%) along the radio-tail, suggesting that the interaction with its surrounding intracluster medium (ICM) is not highly turbulent. With the present data, thermal particle densities at the locations of these sources are estimated to be n > $10^{-5}\;cm^{-3}$.

• 한국정밀공학회지
• /
• 제25권7호
• /
• pp.27-32
• /
• 2008

• 한국광학회:학술대회논문집
• /
• 한국광학회 1989년도 제4회 파동 및 레이저 학술발표회 4th Conference on Waves and lasers 논문집 - 한국광학회
• /
• pp.36-38
• /
• 1989

We have developed Magnetic Pulse Compression System to realize repetitive excimer laser excitation. The principle of this system is to use the large change in permiability owing to the nolinear characteristics of ferro-magnetic material (Metglas2605s-2 metal ribon). Prior to the laser operation, the MPC system was tested with a dummy load (5$\Omega$) and laser head. Laser head has a discharge volume of 1.0 (w) x 2.0 (h) x 20.0(1) cm. This MPC system compressed a 6.2us (FWHM), 80 A pulse into a 0.4us(FWHM), 1.3kA pulse.

• 항공우주기술
• /
• 제7권2호
• /
• pp.76-81
• /
• 2008

우주환경시험팀에서는 발사환경 시 진동 시험을 위해 진동 테스트 장비를 운영 중에 있으며 이를 이용하여 여러 위성 프로그램의 환경시험을 성공적으로 수행하였다. 발사체 및 대형위성의 진동시험을 위한 대형 전자기 시스템의 요구를 충족시키기 위해, 직경 3m급 헤드익스펜더를 가진 대형 멀티 전자기 가진 시스템을 개발하고 있다. 개발 시스템은 160 kN의 추력을 가진 3대의 가진 시스템으로 구성되어 있으며 최대 8톤급 시험체의 진동시험과 300 kNm의 횡 하중 지지할 수 있도록 설계되었다. 본 연구에서는 대형 멀티 전자기 가진 시스템 개발 과정에서 요구되는 각 요소의 설계에 대한 연구 결과를 소개하고자 한다.

• 한국기계가공학회지
• /
• 제14권6호
• /
• pp.77-83
• /
• 2015

This paper is a study of the functions required for the system to quantify the closest target weight by combining several random weights such as chips, snacks, fruits, and vegetables. The multi-head weigher is designed for high-performance applications requiring increased production rates and tight accuracy tolerances. This combination system has 12 heads considered in the form of a rectangular array of $2{\times}6$ or $3{\times}4$. Channel combination can usually occur between 1 and n, and the frequency was the highest with two or three combinations. Experimental result of a combination system for a total target weight was measured at the range from 100g to 500g by increments of 50g, and the average success rate was about 70%. The average elapsed time was about 1.7 seconds, which means it can be used for the packaging of agricultural products with a variety of items.

• 한국군사과학기술학회지
• /
• 제14권5호
• /
• pp.880-887
• /
• 2011

The Wireless Mesh Network(WMN) technology is able to provide an infrastructure for isolated islands, in which it is difficult to install cables or wide area such as battlefield. Therefore, WMN is frequently used to satisfy needs for internet connection and active studies and research on them are in progress. However, as a result of increase in number of hops under hop-by-hop communication environment has caused a significant decrease in throughput and an increase in delay. Considering the heavy traffic of real-time data, such as voice or moving pictures to adaptive WMN, in a military environment. Such phenomenon might cause an issue in fairness index. In order to resolve this issue, we proposed a Cluster Based Multi-channel Assignment Scheme(CB-MAS) for adaptive tactical wireless mesh network. In the CB-MAS, the communication between the Cluster-Head(CH) and cluster number nodes uses a channel has no effect on channels being used by the inter-CH links. Therefore, the CB-MAS can minimize the interference within multi-channel environments. Our Simulation results showed that CB-MAS achieves improved the throughput and fairness index in WMN.