As part of South Korea's national space program, an ongoing initiative involves the development of a satellite constellation, NEONSAT (New-Space Earth Observation Satellite Constellation for National Safety), comprising 11 microsatellites equipped with high-resolution optical payloads. The first prototype of this constellation is scheduled for launch in March this year, with the remaining 10 satellites to be developed and launched sequentially until 2027. To enhance the utility of NEONSAT for national land monitoring and public safety, an application system has been developed to provide easy access to satellite imagery data and analysis processing services. This paper explains the development framework of this land application system to accelerate the generation of analysis products with efficiency, using automated processing techniques on a large number of high-resolution multispectral images acquired by NEONSAT.

While contrastive learning has been the mainstream in self-supervised learning, negative sampling for learning is a challenging issue. To resolve this problem, recent studies instead used a pair of augmented images generated from each image as positive samples and trained models by maximizing the similarity between representation vectors. However, training models using positive samples may not be enough to let the models learn rich representations because these methods cannot consider the information from different images. To address the issue, we propose a learning method that uses clustered positive samples from different images to allow the network to learn better. As the first step, our method learns invariance representations from augmented images. Next, the resulting representation vectors are clustered to learn rich representations. According to our evaluation, it was found that our method achieved better performance than other models using a clustering method from scratch. Moreover, in downstream tasks, our model achieved a higher classification accuracy than Barlow Twins under the same experimental setting. Therefore, adding the distance information on the class centroid to the loss function of self-supervised learning can improve performance by exploiting rich representations in the data.

The increasing deployment of microsatellites in recent years has emerged as a cost-effective and efficient alternative to traditional high-weight satellites. These smaller satellites offer the advantage of frequent revisits, allowing for the monitoring and analysis of dynamic events and changes over time. This study specifically focuses on harnessing the potential of microsatellite imagery in disaster response scenarios, with a particular emphasis on the detection and assessment of wildfires. The research explores the development of methods for the rapid and effective detection of wildfires using microsatellite data, employing the Morphological Active Contours without Edge algorithm for analysis. Additionally, the study introduces the NEONSAT satellite constellation system, a significant initiative aimed at providing efficient analysis results for disaster response efforts. Simulated datasets were utilized to test the thresholding methods, and the evaluation of results was conducted using the Normalized Burned Ratio metric. Future research will focus on validating with actual microsatellite data and comprehensively assessing the performance of different algorithms for disaster response applications.

Diet planning is a basic and regular human activity. Previous studies have considered diet planning a combinatorial optimization problem to generate solutions that satisfy a diet's nutritional requirements. However, this approach does not consider the composition of diets, which is critical for diet recipients' to accept and enjoy menus with high nutritional quality. Without this consideration, feasible solutions for diet planning could not be provided in practice. This suggests the necessity of diet planning with machine learning, which extracts implicit composition patterns from real diet data and applies these patterns when generating diets. This work is original research that defines diet planning as a machine learning problem; we describe diets as sequence data and solve a controllable sequence generation problem. Specifically, we develop the Teacher-forced REINFORCE algorithm to connect neural machine translation and reinforcement learning for composition compliance with nutrition enhancement in diet generation. Through a real-world application to diet planning for children, we validated the superiority of our work over the traditional combinatorial optimization and modern machine learning approaches, as well as human (i.e., professional dietitians) performance. In addition, we construct and open the databases of menus and diets to motivate and promote further research and development of diet planning with machine learning. We believe this work with data science will contribute to solving economic and social problems associated with diet planning.

For the application of agricultural area, remote sensing techniques were studied and applied for its advantages for continuous and quantitative monitoring. Especially, hyperspectral images have been studied for the precise agriculture since they provide chemical and physical information of vegetation. In this study, we analyzed crop types using hyperspectral image data collected by a ground scanner. Spectral mixture analysis, which is widely used for processing hyperspectral images, was adopted for the crop discrimination. Endmember extraction algorithms used in this study were N-FINDR, Vertex Component Analysis (VCA), and Simplex Identification via variable Splitting and Augmented Lagrangian (SISAL), and classification was processed using fully constrained linear spectral unmixing (FCLSU). This study presents the application of spectral mixture analysis for hyperspectral scanner data at canopy level and optimal endmember extraction algorithms for different crop types for precise agriculture.

In this paper we present DenTeach, a smart brush with a doll-shaped tangible device. We designed and to foster children's good tooth brushing habits by explicitly showing the effect of their tooth brushing on the teeth of a physical avatar in real time. A pilot study conducted with 15 children (ages between 6 and 12) shows that there was possibly positive change of their attitudes to tooth brushing and motivations for tooth brushing when they brushed their teeth with DenTeach. As future work, we plan to develop a smartphone app that interacts with DenTeach so that the children users can constantly check their tooth brushing habits with more fun.

Spatial ability is one of mental ability which grounds children's educational development critically, and is used in whole dimensions of our life in conjunction with STEM (Science, Technology, Engineering, and Mathematics) learning. In this paper, we present a gamification tool socalled "Hide and Seek" which nourishes spatial ability. A child seeks a place where a tagger is hidden by comparing Lego village and the corresponding picture taken by the tagger. A concept of this game is comparing the 2-dimensional flat surface of the picture and 3-dimensional Lego village. We conducted user tests of Hide and Seek with children. The aims of user test are observing usability, usage of spatial language and game elements while children are playing the game. From the test, we observed spatial cognition factor and dialogues used during problem solving and usability. Based on the results, we also derived additional concept refinements of the presented educational game system.