• Korean Journal of Remote Sensing
• /
• v.38 no.6_2
• /
• pp.1743-1747
• /
• 2022

Sentinel-2 can be used as proxy data for the Korean Compact Advanced Satellite 500-4 (CAS500-4), also known as Agriculture and Forestry Satellite, in terms of spectral wavelengths and spatial resolution. This letter examined cloud detection for later use in the CAS500-4 based on deep learning technologies. DeepLabV3+, a traditional Convolutional Neural Network (CNN) model, and Shifted Windows (Swin) Transformer, a state-of-the-art (SOTA) Transformer model, were compared using 22,728 images provided by Radiant Earth Foundation (REF). Swin Transformer showed a better performance with a precision of 0.886 and a recall of 0.875, which is a balanced result, unbiased between over- and under-estimation. Deep learning-based cloud detection is expected to be a future operational module for CAS500-4 through optimization for the Korean Peninsula.

• Journal of Korean Elementary Science Education
• /
• v.40 no.3
• /
• pp.390-406
• /
• 2021

This study investigated how elementary students with different spatial ability constructed spatial thinking process about on volcanoes and earthquakes with GeoMapApp-based materials. Students' spatial thinking process was analyzed in terms of spatial concept recognized, tools of spatial representation, and their spatial reasoning to construct topographic structure. The student group with high-scored spatial ability showed the spatial reasoning based on internal representation of building mental images through sectional division of horizontal distance, directly connected with spatial concept, or distorting spatial concept. The student group with low-scored spatial ability built the spatial reasoning directly connected with spatial concept instead of transforming into internal representation, and partially recognized spatial concept on either distance or depth. Based on the results, we argued identifying spatial concepts such as distance, height, or depth from the GeoMapApp data would be funda- mental for the better spatial thinking.

• ETRI Journal
• /
• v.43 no.2
• /
• pp.272-287
• /
• 2021

This article presents a multi-dimensional spatial pattern analysis of crime events in San Francisco. Our analysis includes the impact of spatial resolution on hotspot identification, temporal effects in crime spatial patterns, and relationships between various crime categories. In this work, crime prediction is viewed as a classification problem. When predictions for a particular category are made, a binary classification-based model is framed, and when all categories are considered for analysis, a multiclass model is formulated. The proposed crime-prediction model (HotBlock) utilizes spatiotemporal analysis for predicting crime in a fixed spatial region over a period of time. It is robust under variation of model parameters. HotBlock's results are compared with baseline real-world crime datasets. It is found that the proposed model outperforms the standard DeepCrime model in most cases.

• Journal of Institute of Control, Robotics and Systems
• /
• v.21 no.6
• /
• pp.565-570
• /
• 2015

Supervised machine learning has become popular in discovering context descriptions from sensor data. However, collecting a large amount of labeled training data in order to guarantee good performance requires a great deal of expense and time. For this reason, semi-supervised learning has recently been developed due to its superior performance despite using only a small number of labeled data. In the existing semi-supervised learning algorithms, unlabeled data are used to build a graph Laplacian in order to represent an intrinsic data geometry. In this paper, we represent the unlabeled data as the spatial-temporal dataset by considering smoothly moving objects over time and space. The developed algorithm is evaluated for position estimation of a smartphone-based robot. In comparison with other state-of-art semi-supervised learning, our algorithm performs more accurate location estimates.

• Sleep Medicine and Psychophysiology
• /
• v.7 no.1
• /
• pp.10-17
• /
• 2000

To investigate the neurobiological bases of learning and memory is one of the ambitious goals of modern neuroscience. The progress in this field of recent years has not only brought us closer to understanding the molecular mechanism underlying long-lasting changes in synaptic strength, but it has also provided further evidence that these mechanisms are required for memory formation. Since twenty years ago, several studies for the tests of the hypothesis that NMDA-dependent hippocampal long-term potentiation(LTP) underlies learning have been reported. Also, in the recent year, data from mutant mice showed that a potential role for NMDA-dependent LTP in hippocampal CA1 and spatial learning. Although the current evidence for the role of NMDA receptor in learning and memory is not still obvious, NMDA receptor seems to act as a critical switch for activation of a cascade of events that underlie synaptic plasticity.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.17 no.3
• /
• pp.861-880
• /
• 2023

Semantic segmentation of road scene is the key technology of autonomous driving, and the improvement of convolutional neural network architecture promotes the improvement of model segmentation performance. The existing convolutional neural network has the simplification of learning knowledge and the complexity of the model. To address this issue, we proposed a road scene semantic segmentation algorithm based on multi-task collaborative learning. Firstly, a depthwise separable convolution atrous spatial pyramid pooling is proposed to reduce model complexity. Secondly, a collaborative learning framework is proposed involved with saliency detection, and the joint loss function is defined using homoscedastic uncertainty to meet the new learning model. Experiments are conducted on the road and nature scenes datasets. The proposed method achieves 70.94% and 64.90% mIoU on Cityscapes and PASCAL VOC 2012 datasets, respectively. Qualitatively, Compared to methods with excellent performance, the method proposed in this paper has significant advantages in the segmentation of fine targets and boundaries.

• International Journal of Advanced Culture Technology
• /
• v.11 no.2
• /
• pp.56-63
• /
• 2023

I proposed the architectural guideline for educational environment based on design thinking approach to integrate and enhance learners' activities and achievements. The physical environment of design education learning space should be applied by teaching methods and learning activities, especially for STEAM-based convergent education, the architectural space conditions should support the design process based on design thinking. The learning environment conditions influence design education with physical design factors and learners' communication, and the flexible environment based on design thinking, which is crucial for design education. The 3 steps of design thinking experiences also allow students to learn the context of ideas, skills and outcomes. Therefore, I argued that the learning surrounding based on design thinking needs flexible and mobile, connected, integrated, organized, and team-focused environments to support learners' understanding, participation, and collaboration, and to achieve the design process based on research findings. For spaces for convergent learning environments based on design thinking, common design principles should be reviewed, such as coexistence with technology, safety and security, transparency and spatial extension, multi-purpose space and outdoor learning.

• Journal of the Korean association of regional geographers
• /
• v.8 no.4
• /
• pp.529-541
• /
• 2002

This paper aims to provide a context-specific interpretation on the shifting process of in-house R&D spaces in a large Korean firm in the context of rapidly changing markets and technology. Drawing on the case study of LG Electronics Company, one of the Korea's flagship companies, I examine the causes and mechanisms leading to a shift in domestic R&D spaces and the nature of learning processes between R&D teams and between R&D and other organizational units, particularly manufacturing. It appears that the current reshaping processes of domestic R&D spaces in LGE focus more on the clustering of core R&D laboratories than the geographical integration of conception and execution. However, it should not simply be viewed that such a move would be reduced to the linear model of innovation and organizational learning. Instead, it involves the firm-specific mode of regulating organizational competences. As contextual variables to induce such a firm-specific mode of organizational change, I consider the spatial form of organization, the spatial sources of knowledge and learning, and the powers of relational learning that can be made between distanciated actors and teams.

• Journal of the Korean Geographical Society
• /
• v.39 no.6 s.105
• /
• pp.971-983
• /
• 2004

It is increasingly clear that student mastery of concepts and skills in geographic education is based on a complex set of variables. Sex and gender are the key variables. Much has been written about biological sex differences in learning, but less attention has been paid to the impacts of socio-cultural gender on learning geography. As such, the aims of this paper are two-fold. First, to examine theories which seek to explain why males and females might differ in their geographic and spatial knowledge or skill. Second, to examine the extent of sex differences and gender traits in the geographic learning. The results of study illustrate clearly that there are more similarities than differences between the sexes. Therefore, there are significant gender differences between the preferences of regions, contents, activities in the secondary geographic learning. The results also provide insights into improving contents and method of geographic education.

• The Mathematical Education
• /
• v.52 no.4
• /
• pp.531-547
• /
• 2013

The purpose of the study is to examine how effects of activities using Google SketchUp on students' spatial ability and 3D geometric thinking in measuring the volume of solid figures. By comparing the results from pre- and post-tests between the experimental group and control group, we found that activities using Google SketchUp help students improve their spatial ability in the spatial orientation and visualization. In addition, more than half students in the experimental group moved from level 4 up to level 7 in thinking process of measuring the volume in terms of Battista(2004)'s levels. This study suggests that the instruction with Google SketchUp can help to improve students' spatial ability and 3D geometric thinking in the regular class in middle school. In addition, SketchUp can be an advanced technological tool to support students' self-directed learning, which create an efficient educational environment and a great opportunity to learn geometry in an effective manner.