• Journal of the Korean Society of Radiology
• /
• v.18 no.5
• /
• pp.557-566
• /
• 2024

Ultrasound-guided regional anesthesia is one of the most common techniques used in peripheral nerve blockade by enhancing pain control and recovery time. However, accurate Brachial Plexus (BP) nerve detection and identification remains a challenging task due to the difficulty in data acquisition such as speckle and Doppler artifacts even for experienced anesthesiologists. To mitigate the issue, we introduce a BP nerve small target segmentation network by incorporating BP object detection and U-Net based semantic segmentation into a single deep learning framework based on the multi-scale approach. To this end, the current BP detection and identification was estimated: 1) A RetinaNet model was used to roughly locate the BP nerve region using multi-scale based feature representations, and 2) U-Net was then used by feeding plural BP nerve features for each scale. The experimental results demonstrate that our proposed model produces high quality BP segmentation by increasing the accuracies of the BP nerve identification with the assistance of roughly locating the BP nerve area compared to competing methods such as segmentation-only models.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.49 no.2
• /
• pp.94-101
• /
• 2012

This paper proposes an improved image steganalysis scheme to raise the detection rate of stego images out of cover images. To improve the detection rate of stego image in the steganalysis, tiny variation caused by data hiding should be amplified. For this, we extract feature vectors of cover image and stego image by two steps. First, we separate image into upper 4 bit subimage and lower 4 bit subimage. As a result, stego noise is expanded more than two times. We decompose separated subimages into twelve subbands by applying 3-level Haar wavelet transform and calculate co-occurrence probabilities of two different subbands in the same scale. Since co-occurrence probability of the two wavelet subbands is affected by data hiding, it can be used as a feature to differentiate cover images and stego images. The extracted feature vectors are used as the input to the multilayer perceptron(MLP) classifier to distinguish between cover and stego images. We test the performance of the proposed scheme over various embedding rates by the LSB, S-tool, COX's SS, and F5 embedding method. The proposed scheme outperforms the previous schemes in detection rate to existence of hidden message as well as exactness of discrimination.

• Journal of the Korean earth science society
• /
• v.43 no.1
• /
• pp.140-150
• /
• 2022

This paper focuses on introducing the concept of the multi-scale study on the Ulleungdo-Dokdo Volcanic Group in the East Sea and recent new findings from it. Multi-channel seismic reflection data reveals that the major volcanic activities of the Ulleungdo-Dokdo Volcanic Group took place between 5 and 2.5 Ma, which were propagated from Isabu Tablemount on the eastern end to the Ulleung Island on the western end. The terrestrial Ulleung Island was built via 5 stages, which eventually formed a 3 km-wide caldera, named Nari Caldera, and a volcanic dome, named Albong, within the caldera. The Albong and the unit N-1, the earliest phreatomagmatic explosive phase of the Albong volcano, were generated from a new magma injected into the existing phonolitic body. The generally trachyandesitic bulk rock composition of the pumice in unit N-1 and Albong is attributed to the contamination of the new magma by mafic cumulates at the base of the existing phonolitic chamber. The lines of evidence of a new magma injection point toward that Ulleung Island is an active volcano with a live subvolcanic magma plumbing system.

• Journal of the Korea Society of Computer and Information
• /
• v.17 no.6
• /
• pp.93-100
• /
• 2012

In this paper we have proposed a linear time algorithm for solving the minimum sharable-bandwidth tree construction problem. The public-shared network is a user generated infrastructure on which a user can access the Internet and transfer data from any place via access points with sharable bandwidth. Recently, the idea of constructing the SVC video streaming delivery system on public-shared network has been proposed. To send video stream from the stream server to clients on public-shared network, a tree structure is constructed. The problem of constructing a tree structure to serve the video streaming requests by using minimum amount of sharable bandwidth has been shown to be NP-hard. The previously published algorithms for solving this problem are either unable to find solutions frequently or less efficient. The experimental results showed that our algorithm is excellent both in the success rate of finding solutions and in the quality of solutions.

• Journal of Digital Policy
• /
• v.2 no.3
• /
• pp.1-7
• /
• 2023

In this paper, I proposes a method for performing single image super resolution by separating texture-spatial domains and then classifying features based on detailed information. In CNN (Convolutional Neural Network) based super resolution, the complex procedures and generation of redundant feature information in feature estimation process for enhancing details can lead to quality degradation in super resolution. The proposed method reduced procedural complexity and minimizes generation of redundant feature information by splitting input image into two channels: texture and spatial. In texture channel, a feature refinement process with step-wise skip connections is applied for detail restoration, while in spatial channel, a method is introduced to preserve the structural features of the image. Experimental results using proposed method demonstrate improved performance in terms of PSNR and SSIM evaluations compared to existing super resolution methods, confirmed the enhancement in quality.

• Proceedings of the Korean Society of Computer Information Conference
• /
• 2023.07a
• /
• pp.541-544
• /
• 2023

LLM(Largescale Language Model)의 성능 향상을 위한 비용 효율적인 방법으로 ChatGPT, GPT-4와 같은 초거대 모델의 output에 대해 SLM(Small Language Model)을 finetune하는 방법이 주목받고 있다. 그러나, 이러한 접근법은 주로 범용적인 지시사항 모델을 위한 학습 방법으로 사용되며, 제한된 특정 도메인에서는 추가적인 성능 개선의 여지가 있다. 본 연구는 특정 도메인(Writing Assistant)에서의 성능 향상을 위한 새로운 방법인 Self-Guided Approach를 제안한다. Self-Guided Approach는 (1) LLM을 활용해 시드 데이터에 대해 도메인 특화된 metric(유용성, 관련성, 정확성, 세부사항의 수준별) 점수를 매기고, (2) 점수가 매겨진 데이터와 점수가 매겨지지 않은 데이터를 모두 활용하여 supervised 방식으로 SLM을 미세 조정한다. Vicuna에서 제안된 평가 방법인, GPT-4를 활용한 자동평가 프레임워크를 사용하여 Self-Guided Approach로 학습된 SLM의 성능을 평가하였다. 평가 결과 Self-Guided Approach가 Self-instruct, alpaca와 같이, 생성된 instruction 데이터에 튜닝하는 기존의 훈련 방법에 비해 성능이 향상됨을 확인했다. 다양한 스케일의 한국어 오픈 소스 LLM(Polyglot1.3B, PolyGlot3.8B, PolyGlot5.8B)에 대해서 Self-Guided Approach를 활용한 성능 개선을 확인했다. 평가는 GPT-4를 활용한 자동 평가를 진행했으며, Korean Novel Generation 도메인의 경우, 테스트 셋에서 4.547점에서 6.286점의 성능 향상이 발생했으며, Korean scenario Genration 도메인의 경우, 테스트 셋에서 4.038점에서 5.795 점의 성능 향상이 발생했으며, 다른 유사 도메인들에서도 비슷한 점수 향상을 확인했다. Self-Guided Approach의 활용을 통해 특정 도메인(Writing Assistant)에서의 SLM의 성능 개선 가능성을 확인했으며 이는 LLM에 비용부담을 크게 줄이면서도 제한된 도메인에서 성능을 유지하며, LLM을 활용한 응용 서비스에 있어 실질적인 도움을 제공할 수 있을 것으로 기대된다.

• Proceedings of the Korean Vacuum Society Conference
• /
• 2012.08a
• /
• pp.349-349
• /
• 2012

산화제를 이용 기상중합법을 통해 합성되는 고전도도 Poly (3,4-Ethylenedioxythiophene)(PEDOT) 박막은 OTFT, RFID tag, 또는 연성 디스플레이 같은 분야에 다양한 응용 가능성을 가지고 있으며 이로 인해 최근에 연구가 활발히 진행되고 있다. PEDOT박막의 전극소재로써 가능성은 박막의 중합 정도와 표면 형상에 크게 좌우된다. 특히, Si-웨이퍼 기판 위에 산화제의 균일한 도포 및 산화제 자체의 높은 산도 ($pH{\leq}2$)에 따른 부반응의 억제는 기상중합법을 이용한 PEDOT박막의 합성에 있어 매우 중요하다. PEDOT의 효율적인 중합과 균일한 성장을 위해 산화제에 DUDO 와 PEG-PPG-PEG를 첨가한 혼합 산화제 용액을 제조 기상중합 방법을 통해 PEDOT박막을 제작하였다. 그 결과 산화제만을 사용하여 제작된 박막에 비해 전도도가 최대 3,660 S/cm로 향상된 PEDOT 박막이 합성되었다. 이러한 결과는 PEG-PPG-PEG가 산화제 용액의 균일 도포를 향상시키고 Base Inhibitor로 작용하는 DUDO는 PEDOT 성장 시 중합속도를 조절하고 부반응을 최소화 하여 효율적인 공액 이중 결합의 생성을 촉진한데 주로 기인한다. 따라서 그로인해 조밀하며 마이크로 스케일의 기공이 최소화된 PEDOT박막의 합성이 가능하였다. PEDOT박막의 특성 평가에는 4-point probe, optical microscopy, Field Emission-Scanning Electron Microscope, 등이 사용되었으며 또한 전도도의 향상 원인을 분석하고자 ATR-IR Spectrophotometer를 이용하여 합성된 박막의 작용기를 분석하였다. 이러한 고전도도의 PEDOT 박막이 OTFT의 전극소재로 사용된다면 OTFT소자의 성능 향상에 크게 기여 할 것으로 기대된다.

• Journal of the Economic Geographical Society of Korea
• /
• v.17 no.2
• /
• pp.420-436
• /
• 2014

Regional communities have recently tried to change themselves in order to cope with rapidly changing circumstances. Giving attention to their strategic responses as well as consequences as a result of their countermeasures, this study attempts to suggest a theoretical framework to analyze the process of locality reconstruction in a dynamic and comprehensive way. To achieve this purpose, after figuring out the previous studies' methodological and theoretical limitations, we suggest that locality should be understood as a outcome of the mutual interactions between local, state, and global scale networks.

• Journal of the Korean Vacuum Society
• /
• v.15 no.1
• /
• pp.73-80
• /
• 2006

Thermally-driven effusion of inert gases out from Si(100), into which energetic $\~l keV\;He^+,\;Ne^+,\;A^r+,\;and\;Kr^+ ions$ had been implanted at a moderate substrate temperatures of $\~400 K$, was investigated by means of temperature-programmed desorption (TPD) mass spectrometry. While He effused out broadly over $500\~1,100 K$, Ne, Ar, and Kr effusion occurred sharply at 810, 860, and 875 K, respectively. Hydrogen adsorption/desorption analysis for the ion-treated Si(100) surfaces indicated minimal to severe damage by ions with increasing mass from He to Kr. Implications of these results in light of literature reports are discussed.

• Journal of Korea Multimedia Society
• /
• v.7 no.9
• /
• pp.1321-1328
• /
• 2004

In this article, we present a new DICOM object able to create image-based annotations in DICOM image. Since the proposed image-based annotation uses images themselves of annotation, various types like character, sketch, and scanning image, etc., can be imported into annotation easily. The proposed annotation is inserted into DICOM image directly but they do not influence original DICOM image quality by using independent data channel. The proposed annotation is expected to be very useful to medium and small clinics that cannot afford picture archiving and communication systems or electronic medical record.