• Structural Engineering and Mechanics
• /
• v.56 no.2
• /
• pp.293-315
• /
• 2015

While most of researches on system identification of building structures are aimed at finding modal parameters first and identifying the corresponding physical parameters by using the transformation in terms of transfer functions and cross spectra, etc., direct physical parameter system identification methods have been proposed recently. Due to the problem of signal/noise (SN) ratios, the previous methods are restricted mostly to earthquake records or forced vibration data. In this paper, a theoretical investigation is performed on the influence of wind disturbances on stiffness identification of building structures using micro-tremor at limited floors. It is concluded that the influence of wind disturbances on stiffness identification of building structures using micro-tremor at limited floors is restricted in case of using time-series data for low-rise buildings and does not cause serious problems.

• Coupled systems mechanics
• /
• v.3 no.4
• /
• pp.345-365
• /
• 2014

A significant problem met in engineering practice when designing cable-stayed bridges is the failure of cables. Many different factors can lead to sudden failure of cables, such as corrosion, continuous friction or abrasion, progressive and extended crevice created by fatigue and finally an explosion caused by sabotage or accident, are some of the causes that can lead to the sudden failure of one or more cables. This paper deals with the sudden failure of cables in a special form of cable-stayed bridges with a single line of cables anchored at the central axis of the deck's cross-section. The analysis is carried out by the modal superposition technique where an analytical method developed by the authors in a previous work has been employed.

• Journal of Information Processing Systems
• /
• v.19 no.1
• /
• pp.80-88
• /
• 2023

RGB-thermal (RGB-T) tracking using unmanned aerial vehicles (UAVs) involves challenges with regards to the similarity of objects, occlusion, fast motion, and motion blur, among other issues. In this study, we propose dynamic tracking aggregation (DTA) as a unified framework to perform object detection and data association. The proposed approach obtains fused features based a transformer model and an L1-norm strategy. To link the current frame with recent information, a dynamically updated embedding called dynamic tracking identification (DTID) is used to model the iterative tracking process. For object association, we designed a long short-term tracking aggregation module for dynamic feature propagation to match spatial and temporal embeddings. DTA achieved a highly competitive performance in an experimental evaluation on public benchmark datasets.

• Annual Conference on Human and Language Technology
• /
• 2019.10a
• /
• pp.155-160
• /
• 2019

본 연구에서는 사진 검색을 위한 Attentive Aggregation(주의적 종합) 기반의 언어-시각 크로스모달 임베딩 모델을 제안한다. 본 연구에서는 크로스모달 임베딩을 활용한 검색 과제에서 검색 대상의 임베딩을 계산하는 새로운 방법으로 '질의 기반 종합 검색 대상 임베딩'을 제안하며, Attentive Aggregation 레이어를 활용하여 이를 적용한 크로스모달 임베딩 모델을 제안한다. 제안 모델은 정보량이 많은 사진 데이터로부터 여러 특징을 추출한 뒤 주어진 질의에 따라 이들을 선택적으로 반영한 임베딩을 계산할 수 있으며, 이에 따라 Recall@10 약 0.23, MAP@10 약 0.11, MRR 약 0.13으로 Baseline과 비교하였을 때 크게 향상된 사진 검색 성능을 보였다.

• Annual Conference of KIPS
• /
• 2023.05a
• /
• pp.660-662
• /
• 2023

3차원 포인트 클라우드의 의미적 분할은 환경을 구성하는 물체 단위로 포인트 클라우드를 분할하는 작업으로서, 환경의 3차원적 구성을 이해하고 환경과 상호작용에 필수적인 시각 지능을 요구한다. 본 논문에서는 포인트 클라우드에서 추출하는 3차원 기하학적 특징과 함께 멀티-뷰 영상에서 추출하는 2차원 시각적 특징들도 활용하는 새로운 3차원 포인트 클라우드 의미적 분할 모델 MFNet을 제안한다. 제안 모델은 서로 이질적인 2차원 시각적 특징과 3차원 기하학적 특징의 효과적인 융합을 위해, 새로운 중기 융합 전략과 멀티-모달 교차 주의집중을 이용한다. 본 논문에서는 ScanNetV2 벤치마크 데이터 집합을 이용한 다양한 실험들을 통해, 제안 모델 MFNet의 우수성을 입증한다.

• Journal of Institute of Control, Robotics and Systems
• /
• v.7 no.10
• /
• pp.812-818
• /
• 2001

This paper proposes an optimum design method of structural and control systems, taking a 2-D truss structure as an example. The structure is supposed to be subjected to initial static loads and disturbances. For the structure, a FEM model is formed, and using modal transformation, the equation of motion is transformed into that of modal coordinates in order to reduce the D.O.F. of the FEM model. The structure is controlled by an output feedback $H^$\infty$$ controller to suppress the effect of the disturbances. The design variables of the simultaneous optimal design of control-structure systems are the cross sectional areas of truss members. The structural objective function is the structural weight. The control objective function is the $H^$\infty$$ norm, that is, the performance index of control. The second structural objective function is the energy of the response related to the initial state, which is derived from the time integration of the quadratic form of the state in the closed-loop system. In a numerical example, simulations have been carried out. Through the consideration of structural weight and $H^$\infty$$ norm, an advantage of the simultaneous optimum design of structural and control systems is shown. Moreover, while the optimized performance index of control is almost kept, we can acquire better design of structural strength.

• The Journal of the Acoustical Society of Korea
• /
• v.37 no.4
• /
• pp.181-187
• /
• 2018

In this paper, sound absorption of micro-perforated elastic plates installed in an impedance tube of a circular cross-section is discussed using an analytic method. Vibration of the plates and sound pressure fields inside the duct are expressed in terms of an infinite series of modal functions, where modal functions in the radial direction is given in terms of the Bessel functions. Under the plane wave assumption, a low frequency approximation is derived by including the first few plate modes, and the sound absorption coefficient is given in terms of an equivalent impedance of a single surface. The sound absorption coefficient using the proposed formula is in excellent agreement with the result by the FEM (Finite Element Method), and shows dips and peaks at the natural frequencies of the plate. When the perforation ratio is very small, the sound absorption coefficient is dominated by the vibration effect. However, when the perforation ratio reaches a certain value, the sound absorption is mainly governed by the rigid MPP (Micro-Perforated Plate), while the vibration effect becomes very small.

• Journal of Information Processing Systems
• /
• v.10 no.1
• /
• pp.1-22
• /
• 2014

Recent technological advances provide the opportunity to use large amounts of multimedia data from a multitude of sensors with different modalities (e.g., video, text) for the detection and characterization of criminal activity. Their integration can compensate for sensor and modality deficiencies by using data from other available sensors and modalities. However, building such an integrated system at the scale of neighborhood and cities is challenging due to the large amount of data to be considered and the need to ensure a short response time to potential criminal activity. In this paper, we present a system that enables multi-modal data collection at scale and automates the detection of events of interest for the surveillance and reconnaissance of criminal activity. The proposed system showcases novel analytical tools that fuse multimedia data streams to automatically detect and identify specific criminal events and activities. More specifically, the system detects and analyzes series of incidents (an incident is an occurrence or artifact relevant to a criminal activity extracted from a single media stream) in the spatiotemporal domain to extract events (actual instances of criminal events) while cross-referencing multimodal media streams and incidents in time and space to provide a comprehensive view to a human operator while avoiding information overload. We present several case studies that demonstrate how the proposed system can provide law enforcement personnel with forensic and real time tools to identify and track potential criminal activity.

• Journal of the Korean Society of Hazard Mitigation
• /
• v.11 no.1
• /
• pp.7-13
• /
• 2011

Accurate response analysis of long span bridges subjected to seismic excitation is important for earthquake hazard mitigation. In this paper, the performance of a typical four span continuous reinforced concrete bridge model subjected to asynchronous multiple seismic excitations at the supports is investigated in both the time and frequency domains and the results are compared with that from a relevant uniform support excitations. In the time domain analysis, a linear modal superposition approach is used to compute the peak response values. In the frequency domain analysis, linear random vibration theory is used to determine the root mean square response values where the cross correlation effects between the modal and the support excitations on the seismic response of the bridge model are included. From the two sets of results, a practical range of peak factors which are defined to be the ratio of peak and the root mean square responses are suggested for displacements and forces in members. With reliable practical values of peak factors, the frequency domain analysis is preferred for the performance based design of bridges because of the computational advantage and the generality of the results as the time domain analysis only yields results for the specific excitation input.

• Science of Emotion and Sensibility
• /
• v.20 no.2
• /
• pp.3-10
• /
• 2017

This research investigated the cross-modal correspondence between the turbidity and taste of beverages. For an empirical study, we employed two types of achromatic beverages: rice wine and a soft drink. The turbidity of each drink varied between 0.0625~32 g/L, and 0.0078~4 g/L, respectively, and the stimuli were prepared in ten levels according to cognitive turbidity. In the study, participants (N=35) rated each drink stimulus using a 3-point Likert scale with regard to five basic tastes: sweet, salty, bitter, sour, and preferred. In addition, six specific tastes were included that deliberately describe rice wine and the soft drink. Three were yeasty, alcoholic, and astringent for the rice wine, and the other three were creamy, bubbly, and syrupy for the soft drink. Based on participants' assessments, the turbidity of rice wine is highly positively correlated with all five basic tastes. In contrast, the turbidity of the soft drink was positively correlated with sour only. Concerning preference, the most preferred turbidity of rice wine was 4.6~20 g/L, which is close to the turbidity of existing products on the market. Furthermore, except for astringent and syrupy, all tastes were influenced by the turbidity, which implies the potential of turbidity as a new visual parameter to communicate the taste experience of beverages.