• Annual Conference on Human and Language Technology
• /
• 2014.10a
• /
• pp.38-43
• /
• 2014

본 논문은, 현존하는 영어 FrameNet 데이터를 기반으로 하여, FrameNet에 대한 전문 지식이 없는 번역가들을 통해 수행할 수 있는 한국어 FrameNet의 수동 구축 개발 과정을 제시한다. 우리 연구팀은 실제로, NLTK가 제공하는 영어 FrameNet 버전 1.5의 Full Text를 이루고 있는 5,945개의 문장들 중에서, Frame 데이터를 가진 4,025개의 문장들을 추출해내어, 번역가들에 의해 한국어로 수동번역 함으로써, 한국어 FrameNet 구축 개발을 향한 의미 있는 초석을 마련하였으며, 제시한 방법의 실효성을 입증하는 연구결과들을 웹에 공개하기도 하였다.

• Annual Conference on Human and Language Technology
• /
• 2020.10a
• /
• pp.227-232
• /
• 2020

본 논문은 의미역 주석(Semantic Role Labeling) 자원인 FrameNet을 준구어 말뭉치인 드라마 대본에 주석하는 과정과 주석 결과에 대해 서술한다. 본 논문에서는 프레임 - 프레임 논항 구조의 주석 범위를 한 문장에서 여러 발화로 이루어진 장면 (Scene) 단위의 대본으로 확장하여 문장 내에서 생략된 프레임 논항(Null-Instantiated Frame Elements)을 장면 단위 대본 내의 다른 발화에서 복원하였다. 본 논문은 프레임 자동 분석기를 통해 동일한 드라마의 한국어, 영어 대본에 FrameNet 주석을 한 드라마 대본을 선발된 주석자에 의해 대상 어휘 적합성 평가, 프레임 적합성 평가, 생략된 프레임 논항 복원을 실시하고, 자동 주석된 대본과 주석자 작업 후의 대본 결과를 비교한 결과와 예시를 제시한다. 주석자가 자동 주석된 대본 중 총 2,641개 주석 (한국어 1,200개, 영어 1,461개)에 대하여 대상 어휘 적합성 평가를 실시하여 한국어 190개 (15.83%), 영어 226개 (15.47%)의 부적합 대상 어휘를 삭제하였다. 프레임 적합성 평가에서는 대상 어휘에 자동 주석된 프레임의 적합성을 평가하여 한국어 622개 (61.68%), 영어 473개 (38.22%)의 어휘에 대하여 새로운 프레임을 부여하였다. 생략된 프레임 논항을 복원한 결과 작업된 평균 프레임 논항 개수가 한국어 0.780개에서 2.519개, 영어 1.290개에서 2.253개로 증가하였다.

• Journal of the Korean Society of Fisheries and Ocean Technology
• /
• v.42 no.1
• /
• pp.1-10
• /
• 2006

Present study was conducted to improve the spring frame net trap for conger eel, Conger myriaster which prevents by-catch and protects immature fish. A series of comparative fishing experiment was carried out in water off the south-east coast of Korea from Nov. 2004 to Jul. 2005. And analyzed the amount of catch, by-catch rate, and CPUE, etc. of experimental traps which were the spring frame net traps, sorted four types of flappers. And the compared gear was a plastic conger eel pot. The conclusions were as follows. By-catch rates of spring frame net trap with the compressed flapper bound by nylon mono-filament was about 0-3%. CPUE of spring frame net trap with the compressed flapper was about 50-60% lower than that of traditional and plastic pot. And in case of conger eel that the total length is over 350mm, CPUE was little different on each type of flapper of every mesh size.

• IEMEK Journal of Embedded Systems and Applications
• /
• v.11 no.4
• /
• pp.219-225
• /
• 2016

Visual object tracking is one of the key tasks in computer vision. Robust trackers should address challenging issues such as fast motion, deformation, occlusion and so on. In this paper, we therefore propose a visual object tracking method that exploits inter-frame correlations of convolutional feature maps in Convolutional Neural Net (ConvNet). The proposed method predicts the location of a target by considering inter-frame spatial correlation between target location proposals in the present frame and its location in the previous frame. The experimental results show that the proposed algorithm outperforms the state-of-the-art work especially in hard-to-track sequences.

• Smart Media Journal
• /
• v.9 no.1
• /
• pp.23-29
• /
• 2020

This paper presents an approach for dynamic hand gesture recognition by using algorithm based on 3D Convolutional Neural Network (3D_CNN), which is later extended to 3D Residual Networks (3D_ResNet), and the neural network based key frame selection. Typically, 3D deep neural network is used to classify gestures from the input of image frames, randomly sampled from a video data. In this work, to improve the classification performance, we employ key frames which represent the overall video, as the input of the classification network. The key frames are extracted by SegNet instead of conventional clustering algorithms for video summarization (VSUMM) which require heavy computation. By using a deep neural network, key frame selection can be performed in a real-time system. Experiments are conducted using 3D convolutional kernels such as 3D_CNN, Inflated 3D_CNN (I3D) and 3D_ResNet for gesture classification. Our algorithm achieved up to 97.8% of classification accuracy on the Cambridge gesture dataset. The experimental results show that the proposed approach is efficient and outperforms existing methods.

• Journal of IKEEE
• /
• v.27 no.4
• /
• pp.531-536
• /
• 2023

This paper proposes a sliding window method with frame feature insertion for immediate attack detection on in-vehicle networks. This method guarantees real-time attack detection by labeling based on the attack status of the current frame. Experiments show that the proposed method improves detection performance by giving more weight to the current frame in CNN computation. The proposed model was designed based on a lightweight LeNet-5 architecture and it achieves 100% detection for DoS attacks. Additionally, by comparing the complexity with conventional models, the proposed model has been proven to be more suitable for resource-constrained devices like ECUs.

• KIPS Transactions on Software and Data Engineering
• /
• v.13 no.1
• /
• pp.17-27
• /
• 2024

Video frame interpolation is an important technique used in the field of video and media, as it increases the continuity of motion and enables smooth playback of videos. In the study of video frame interpolation using deep learning, Kernel Based Method captures local changes well, but has limitations in handling global changes. In this paper, we propose a new U-Net structure that applies feature map differentiation and two directions to focus on capturing major changes to generate intermediate frames more accurately while reducing the number of parameters. Experimental results show that the proposed structure outperforms the existing model by up to 0.3 in PSNR with about 61% fewer parameters on common datasets such as Vimeo, Middle-burry, and a new YouTube dataset. Code is available at https://github.com/Go-MinSeong/SF-AdaCoF.

• Journal of Communications and Networks
• /
• v.1 no.2
• /
• pp.134-142
• /
• 1999

We have developed fully functional Video Server and Client applications which can transmit, receive, decompress and display compressed video over various networks. Our video trans-port allows dynamic rate control feedback, loss detection, and repair requests from Clients to the Server. Our experiments show how feedvack-before-degradation scheme for rate adaptation maintains good display frame-rate for video playback. We show how the playback degradation(reduction in display frame-rate) oc-curs and what happens if corrective measures are not taken to im-prove the situation. The degradation is attributed to the increased internal kernel buffering which consumes scarce CPU resources. We demontrate with our experimental results that ServerNet, with improved hardware delivery guarantees, can significantly reduce host CPU resource consumption while serving video streams. We present the maximum number of streams which can be served for each of ATM and ServerNet interconnects. The appropriate user-level packet size for the video server are also determined for each case.

• Journal of Korean Society of Steel Construction
• /
• v.21 no.1
• /
• pp.63-70
• /
• 2009

Failure modes result in fracture or tearing, which may cause deterioration of resistance and reduction of inelastic deformation capacity. The potential failure modes for Special Concentrically Braced Frames (SCBFs) include fracture or tearing of the brace, net section fracture of the brace or gusset plate, fracture of the gusset plate welds, shear fracture of the bolts, block shear, excessive bolt bearing deformation, and buckling of the gusset plate. HSS tubular braces are commonly used in SCBFs, and net section fracture of the tubular brace may also occur through the brace net section at the end of the slot cut into the tube to slip over the gusset plate. This failure mode is categorized as a tension failure mode, and may cause dramatic loss of resistance and brittle behavior. Net section reinforcement is required according to AISC design specifications (AISC 2001). In this paper, the need to reinforce the net section area was discussed. Initially, the results of the net section fracture tests done by the University of California in Berkeley were presented with the modeling of these tests using FE models. To investigate the possibility of net section fracture in an actual frame, the slot end hole model was adapted to the frame FE model, and alternate near-fault histories were applied with tension-dominated cycles, since previous analyses showed that loading history was the most critical factor in net section fracture. The need for this reinforcement (cover plate) and the tension-dominated near-fault history were investigated.

• Proceedings of the KIEE Conference
• /
• 1999.07b
• /
• pp.838-840
• /
• 1999

In this paper, modeling and analysis of discrete event systems by temporal logic frame works and petri net is considered. The reachability tree of the petri net can be used to solve the safeness, boundedness, conservation and coverability problems of discrete event systems. But the reachability tree of the petri net do not solve reachability and liveness problems in general. We proposed a method that synthesised the petri net and the temporal logic frameworks. This method slove some problems of petri net by logical representation of temporal logic frameworks.