• 한국디지털건축인테리어학회논문집
• /
• 제11권2호
• /
• pp.61-70
• /
• 2011

This study aims to establish the concept about a new approaching method dor remodeling in the planning step. also, this study sets up the way how to approach in a dimensional way by classifying the conceptual composition of the target housing as a method. The dimensional definition is as below. The one dimensional approach: the spot. the two dimensional approach: the line, the three dimensional approach: the apatial mass, the four dimensional approach: space + time, the n dimensional (multi-dimensional) approach: the space + time + emotion. The research findings are as follows : For the purpose of remodeling, the old housing should be viewed in the three dimensional shape and space. It is defined as multi-dimensional module method to design the shape and space of the target housing in a multi-dimensional point, considering the three dimensional space composition, the emotion of the user, the housing itself. and the time of the user.

• 한국디지털건축인테리어학회논문집
• /
• 제5권1호
• /
• pp.1-8
• /
• 2005

The major goal of this study is drawing design plan factors considering residents' life cycle and life style when remodeling design is needed. By drawing such design factors, the study will propose an alternative remodeling design plan. Such design method was defined as Multi-Dimensional module design method. And constructing data-system based on Multi-Dimensional module was secondary goal in this study. The details are as follows: Based on analysis of Residents' demand for remodeling and features of emotion-design applied to remodeling, the system which is suitable for application in remodeling design method was conducted. Based on the type of construction materials, the level value was set up according to emotion-factor application. Also, by conducting coding operation according to the distinction of construction materials, this study seeks the unit drawing method of information for constructing emotion-information system.

• 한국주거학회논문집
• /
• 제16권4호
• /
• pp.121-129
• /
• 2005

The major goal of this study is drawing design plan factors considering residents' life cycle and life style when remodeling design is needed. By drawing such design factors, the study will propose an alternative remodeling design plan. Such design method was defined as Multi-Dimensional module design method. And constructing data-system based on Multi-Dimensional module was secondary goal in this study. The details are as follows: Based on analysis of Residents' demand for remodeling and features of emotion-design applied to remodeling, the system which is suitable for application in remodeling design method was conducted. Based on the type of construction materials, the level value was set up according to emotion-factor application. Also, by conducting coding operation according to the distinction of construction materials, this study seeks the unit drawing method of information for constructing emotion-information system.

• 전자공학회논문지A
• /
• 제32A권8호
• /
• pp.107-118
• /
• 1995

Three-dimensional simulator for thermal oxidation process is developed. The simulator is consisted by two individual module, one is analytic-model module and the other is numerical-model module. The analytic-model which uses simple complementary-error function guarantees fast calculation in prediction of multi-dimensional oxidation process. The numerical-model which is based on boundary element method (BEM), has a good accuracy and suitable for various process conditions. The results of this study show that oxide growth is retarded at the corner of hole structure and enhanced at the corner of island structure. These effects are reson of different distribution of oxidant diffusion and mask stress. The utility of models and simulator developed in this study is demonstrated by using it to predict not only traditional shape of LOCOS but also process effects in small geometry.

• 한국멀티미디어학회논문지
• /
• 제23권12호
• /
• pp.1540-1551
• /
• 2020

Aiming at the problem that the existing human behavior recognition algorithm cannot fully utilize the multi-level spatio-temporal information of the network, a human behavior recognition algorithm based on a dense three-dimensional residual network is proposed. First, the proposed algorithm uses a dense block of three-dimensional residuals as the basic module of the network. The module extracts the hierarchical features of human behavior through densely connected convolutional layers; Secondly, the local feature aggregation adaptive method is used to learn the local dense features of human behavior; Then, the residual connection module is applied to promote the flow of feature information and reduced the difficulty of training; Finally, the multi-layer local feature extraction of the network is realized by cascading multiple three-dimensional residual dense blocks, and use the global feature aggregation adaptive method to learn the features of all network layers to realize human behavior recognition. A large number of experimental results on benchmark datasets KTH show that the recognition rate (top-l accuracy) of the proposed algorithm reaches 93.52%. Compared with the three-dimensional convolutional neural network (C3D) algorithm, it has improved by 3.93 percentage points. The proposed algorithm framework has good robustness and transfer learning ability, and can effectively handle a variety of video behavior recognition tasks.

• 한국컴퓨터산업학회논문지
• /
• 제2권10호
• /
• pp.1291-1300
• /
• 2001

이 논문은 실시간 변화가 가능한 동적 그래픽스를 기반하며, 인간의 감성적 인식을 위해서 이미지 영상을 다차원 데이터의 그래픽 요소로 조작하는 시각화 표현 기법으로 연구되었다. 이 구현의 중요한 사상은 사람의 얼굴 특징 점과 기존의 화상 인식 알고리즘을 바탕으로 획득한 모수 제어 값을 다차원 데이터에 대응시켜 그 변화하는 수축 표정에 따라 감성 표현의 가상 이미지를 생성하는 이미지그래픽으로 표현한다. 제안된 DyFIG 시스템은 감성적인 표현을 할 수 있는 얼굴 그래픽의 모듈을 제안하고 구현하며, 조작과 실험을 통해 감성 데이터 표현 기술과 기법이 실현 가능함을 보인다.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제16권6호
• /
• pp.1833-1848
• /
• 2022

Named entity recognition (NER) is an important basic task in the field of Natural Language Processing (NLP). Recently deep learning approaches by extracting word segmentation or character features have been proved to be effective for Chinese Named Entity Recognition (CNER). However, since this method of extracting features only focuses on extracting some of the features, it lacks textual information mining from multiple perspectives and dimensions, resulting in the model not being able to fully capture semantic features. To tackle this problem, we propose a novel Multi-view Semantic Feature Fusion Model (MSFM). The proposed model mainly consists of two core components, that is, Multi-view Semantic Feature Fusion Embedding Module (MFEM) and Multi-head Self-Attention Mechanism Module (MSAM). Specifically, the MFEM extracts character features, word boundary features, radical features, and pinyin features of Chinese characters. The acquired font shape, font sound, and font meaning features are fused to enhance the semantic information of Chinese characters with different granularities. Moreover, the MSAM is used to capture the dependencies between characters in a multi-dimensional subspace to better understand the semantic features of the context. Extensive experimental results on four benchmark datasets show that our method improves the overall performance of the CNER model.

• Nuclear Engineering and Technology
• /
• 제31권3호
• /
• pp.344-363
• /
• 1999

A multi-dimensional thermal-hydraulic system analysis code, MARS 1.3.1, has been developed in order to have the realistic analysis capability of two-phase thermal-hydraulic transients for pressurized water reactor (PWR) plants. As the backbones for the MARS code, the RELAP5/MOD3.2.1.2 and COBRA-TF codes were adopted in order to take advantages of the very general, versatile features of RELAP5 and the realistic three-dimensional hydrodynamic module of COBRA-TF. In the MARS code, all the functional modules of the two codes were unified into a single code first. Then, the source codes were converted into the standard Fortran 90, and then they were restructured using a modular data structure based on "derived type variables" and a new "dynamic memory allocation" scheme. In addition, the Windows features were implemented to improve user friendliness. This paper presents the developmental work of the MARS version 1.3.1 including the hydrodynamic model unification, the heat structure coupling, the code restructuring and modernization, and their verifications.their verifications.

• Nuclear Engineering and Technology
• /
• 제53권10호
• /
• pp.3182-3195
• /
• 2021

The implementation and validation of multi-dimensional (multi-D) features in thermal-hydraulic system codes aims to extend the application of these codes towards multi-scale simulations. The main goal is the simulation of large-scale three-dimensional effects inside large volumes such as piping or vessel. This novel approach becomes especially relevant during the simulation of accidents with strongly asymmetric flow conditions entailing density gradients. Under such conditions, coolant mixing is a key phenomenon on the eventual variation of the coolant temperature and/or boron concentration at the core inlet and on the extent of a local re-criticality based on the reactivity feedback effects. This approach presents several advantages compared to CFD calculations, mainly concerning the model size and computational efforts. However, the range of applicability and accuracy of the newly implemented physical models at this point is still limited and needs to be further extended. This paper aims at contributing to the validation of the multi-D features of the system code ATHLET based on the simulation of the Tests 1.1 and 2.1, conducted at the test facility ROCOM. Overall, the multi-D features of ATHLET predict reasonably well the evolution from both experiments, despite an observed overprediction of coolant mixing at the vessel during both experiments.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제17권10호
• /
• pp.2768-2787
• /
• 2023

As an important research direction of the application of computer science in the medical field, the automatic generation technology of radiology report has attracted wide attention in the academic community. Because the proportion of normal regions in radiology images is much larger than that of abnormal regions, words describing diseases are often masked by other words, resulting in significant feature loss during the calculation process, which affects the quality of generated reports. In addition, the huge difference between visual features and semantic features causes traditional multi-modal fusion method to fail to generate long narrative structures consisting of multiple sentences, which are required for medical reports. To address these challenges, we propose a multi-feature optimization Transformer (MFOT) for generating radiology reports. In detail, a multi-dimensional mapping attention (MDMA) module is designed to encode the visual grid features from different dimensions to reduce the loss of primary features in the encoding process; a feature pre-fusion (FP) module is constructed to enhance the interaction ability between multi-modal features, so as to generate a reasonably structured radiology report; a detail enhanced attention (DEA) module is proposed to enhance the extraction and utilization of key features and reduce the loss of key features. In conclusion, we evaluate the performance of our proposed model against prevailing mainstream models by utilizing widely-recognized radiology report datasets, namely IU X-Ray and MIMIC-CXR. The experimental outcomes demonstrate that our model achieves SOTA performance on both datasets, compared with the base model, the average improvement of six key indicators is 19.9% and 18.0% respectively. These findings substantiate the efficacy of our model in the domain of automated radiology report generation.