• Proceedings of the Korean Society of Computer Information Conference
• /
• 2022.07a
• /
• pp.99-100
• /
• 2022

본 논문에서는 유방암 초음파 이미지를 학습한 multi-modal 구조를 이용하여 유방암을 판별하는 인공지능을 제안한다. 학습된 인공지능은 유방암을 판별과 동시에, 설명 가능한 인공지능 기법과 ROI를 함께 사용하여 종양의 위치를 나타내준다. 시각적으로 판단 근거를 제시하기 때문에 인공지능의 판단 신뢰도는 더 높아진다.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2023.11a
• /
• pp.526-527
• /
• 2023

Speech emotion recognition(SER) is one of the interesting topics in the machine learning field. By developing multi-modal speech emotion recognition system, we can get numerous benefits. This paper explain about fusing BERT as the text recognizer and CNN as the speech recognizer to built a multi-modal SER system.

• KIPS Transactions on Software and Data Engineering
• /
• v.12 no.12
• /
• pp.505-518
• /
• 2023

3D point cloud semantic segmentation is a computer vision task that involves dividing the point cloud into different objects and regions by predicting the class label of each point. Existing 3D semantic segmentation models have some limitations in performing sufficient fusion of multi-modal features while ensuring both characteristics of 2D visual features extracted from RGB images and 3D geometric features extracted from point cloud. Therefore, in this paper, we propose MMCA-Net, a novel 3D semantic segmentation model using 2D-3D multi-modal features. The proposed model effectively fuses two heterogeneous 2D visual features and 3D geometric features by using an intermediate fusion strategy and a multi-modal cross attention-based fusion operation. Also, the proposed model extracts context-rich 3D geometric features from input point cloud consisting of irregularly distributed points by adopting PTv2 as 3D geometric encoder. In this paper, we conducted both quantitative and qualitative experiments with the benchmark dataset, ScanNetv2 in order to analyze the performance of the proposed model. In terms of the metric mIoU, the proposed model showed a 9.2% performance improvement over the PTv2 model using only 3D geometric features, and a 12.12% performance improvement over the MVPNet model using 2D-3D multi-modal features. As a result, we proved the effectiveness and usefulness of the proposed model.

• Journal of KSNVE
• /
• v.9 no.3
• /
• pp.525-534
• /
• 1999

The present paper proposes a generalized modal analysis procedure for non-uniform, distributed-parameter rotor-bearing systems. An exact element matrix is derived for a Timoshenko shaft model which contains rotary inertia, shear deformation, gyroscopic effect and internal damping. Complex coordinates system is adopted for the convenience in formulation. A generalized orthogonality condition is provided to make the modal decomposition possible. The generalized modal analysis by using a modal decomposition delivers exact and closed form solutions both for frequency and time responses. Two numerical examples are presented for illustrating the proposed method. The numerical study proves that the proposed method is very efficient and useful for the analysis of distributed-parameter rotor-bearing systems.

• Science of Emotion and Sensibility
• /
• v.14 no.4
• /
• pp.627-636
• /
• 2011

The purpose of this study is to develop the multi-modal interaction system. This system provides realistic and an immersive experience through multi-modal interaction. The system recognizes user behavior, intention, and attention, which overcomes the limitations of uni-modal interaction. The multi-modal interaction system is based upon gesture interaction methods, intuitive gesture interaction and attention evaluation technology. The gesture interaction methods were based on the sensors that were selected to analyze the accuracy of the 3-D gesture recognition technology using meta-analysis. The elements of intuitive gesture interaction were reflected through the results of experiments. The attention evaluation technology was developed by the physiological signal analysis. This system is divided into 3 modules; a motion cognitive system, an eye gaze detecting system, and a bio-reaction sensing system. The first module is the motion cognitive system which uses the accelerator sensor and flexible sensors to recognize hand and finger movements of the user. The second module is an eye gaze detecting system that detects pupil movements and reactions. The final module consists of a bio-reaction sensing system or attention evaluating system which tracks cardiovascular and skin temperature reactions. This study will be used for the development of realistic digital entertainment technology.

• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.26 no.3A
• /
• pp.497-512
• /
• 2006

In this paper, a simple but effective analysis procedure to estimate seismic capacities of multi-span continuous bridge structures is proposed on the basis of modal pushover analysis considering all the dynamic modes of structure. Unlike previous studies, the proposed method eliminates the coupling effects induced from the direct application of modal decomposition by introducing an identical stiffness ratio and an approximate elastic deformed shape. Moreover, in addition to these two introductions, the use of an appropriate distributed load {P} makes it possible to predict the dynamic responses for all kinds of bridge structures through a simpler analysis procedure. Finally, in order to establish the validity and applicability of the proposed method, correlation studies between rigorous nonlinear time history analysis and the proposed method are conducted for multi-span continuous bridges.

• Mass Spectrometry Letters
• /
• v.2 no.4
• /
• pp.92-95
• /
• 2011

Pyrophosphates are the key intermediates in the biosynthesis of isoprenoids, and their concentrations could reveal the benefits of statins in cardiovascular diseases. Quantitative analysis of five pyrophosphates, including isopentenyl pyrophosphate (IPP), dimethylallyl pyrophosphate (DMAPP), geranyl pyrophosphate (GPP), farnesyl pyrophosphate (FPP), and geranylgeranyl pyrophosphate (GGPP), was performed using liquid chromatography-tandem mass spectrometry (LC-MS/MS) in negative ionization mode. After dilution with methanol, samples were separated on a 3 ${\mu}m$ particle multi-modal $C_{18}$ column ($50{\times}2$ mm) and quantified within 10 min. The gradient elution consists of 10 mM ammonium bicarbonate and 0.5% triethylamine (TEA) in water and 0.1% TEA in 80% acetonitrile was used at the flow rate of 0.4 mL/min. Overall recoveries were 51.4-106.6%, while the limit of quantification was 0.05 ${\mu}g$/mL for GPP and FPP and 0.1 ${\mu}g$/mL for IPP, DMAPP, and GGPP. The precision (% CV) and accuracy (% bias) of the assay were 1.9-12.3% and 89.6-111.8%, respectively, in 0.05-10 ${\mu}g$/mL calibration ranges ($R^2$ > 0.993). The devised LC-MS/MS technique with the multi-modal $C_{18}$ column can be used to estimate the biological activity of pyrophosphates in plasma and may be applicable to cardiovascular events with cholesterol metabolism as well as the drug efficacy of statins.

• Proceedings of the Korean Society for Noise and Vibration Engineering Conference
• /
• 2002.05a
• /
• pp.441-447
• /
• 2002

KSR-III(Korea Sounding Rocket - III), which is being developed by Space Technology R&D Division of KARI(Korea Aerospace Research Institute) will be launched in late 2002. It is a three-stage, liquid propellant rocket which can reach 250 km altitude and will carry out observation of ozone layer and scientific experiments, such as microgravity experiment, and atmospheric measurement. KSR-III is believed to be an intermediate to the launch vehicle capable of carrying a satellite to its orbit. Space Test Department of KARI performed GVT(Ground Vibration Test) fer KSR-III EM at Rocket Test Building of KARI. GVT is very important for predicting the behavior of rocket in its operation, developing flight control program and performing aerodynamic analysis. This paper gives an introduction of rocket GVT configuration and information on test procedures, techniques and results of It. In this test. to simulate free-free condition, test object hung in the air laterally by 4 bungee cords specially devised. For the excitation of test object, pure random signal by two electromagnetic shakers was used and total 22 frequency response functions were achieved. Polyreference parameter estimation was performed to identify the modal parameters with MIMO(Multi-Input-Multi-Output) method. As the result of the test, low frequency mode shapes and modal parameters below 60Hz were identified

• Proceedings of the Korea Institute of Convergence Signal Processing
• /
• 2006.06a
• /
• pp.57-62
• /
• 2006

휴대용 단말기의 소형화 및 지능화와 더불어 차세대 PC 기반의 유비쿼터스 컴퓨팅에 대한 관심이 높아짐에 따라 최근에는 펜이나 음성 입력 멀티미디어 등 여러 가지 대화 모드를 구비한 멀티 모달 상호작용 (Multi-Modal Interaction MMI)에 대한 연구가 활발히 진행되고 있다. 따라서, 본 논문에서는 잡음 환경에서의 명확한 의사 전달 및 휴대용 단말기에서의 음성-제스처 통합 인식을 위한 인터페이스의 연구를 목적으로 Voice-XML과 Wearable Personal Station(WPS) 기반의 음성 및 내장형 수화 인식기를 통합한 멀티 모달 명령어 인식 시스템 (Multi-Modal Instruction Recognition System : MMIRS)을 제안하고 구현한다. 제안되어진 MMIRS는 한국 표준 수화 (The Korean Standard Sign Language : KSSL)에 상응하는 문장 및 단어 단위의 명령어 인식 모델에 대하여 음성뿐만 아니라 화자의 수화제스처 명령어를 함께 인식하고 사용함에 따라 잡음 환경에서도 규정된 명령어 모델에 대한 인식 성능의 향상을 기대할 수 있다. MMIRS의 인식 성능을 평가하기 위하여, 15인의 피험자가 62개의 문장형 인식 모델과 104개의 단어인식 모델에 대하여 음성과 수화 제스처를 연속적으로 표현하고, 이를 인식함에 있어 개별 명령어 인식기 및 MMIRS의 평균 인식율을 비교하고 분석하였으며 MMIRS는 문장형 명령어 인식모델에 대하여 잡음환경에서는 93.45%, 비잡음환경에서는 95.26%의 평균 인식율을 나타내었다.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.12 no.8
• /
• pp.1456-1462
• /
• 2008

In general, it is known a single biometric-based personal authentication has limitation to improve recognition rate due to weakness of individual recognition scheme. The recognition rate of face recognition system can be reduced by environmental factor such as illumination, while speaker verification system does not perform well with added surrounding noise. In this paper, a multi-modal biometric system composed of face and voice recognition system is proposed in order to improve the performance of the individual authentication system. The proposed empirical weight sum rule based on the reliability of the individual authentication system is applied to improve the performance of multi-modal biometrics. Since the proposed system is implemented using JAVA applet with security function, it can be utilized in the field of user authentication on the generic Web.