• KIPS Transactions on Software and Data Engineering
• /
• v.12 no.12
• /
• pp.519-524
• /
• 2023

In hashing-based image retrieval, the hash code of a manipulated image is different from the original image, making it difficult to search for the same image. This paper proposes and evaluates a self-supervised deephashing model that generates perceptual hash codes from feature information such as texture, shape, and color of images. The comparison models are autoencoder-based variational inference models, but the encoder is designed with a fully connected layer, convolutional neural network, and transformer modules. The proposed model is a variational inference model that includes a SimAM module of extracting geometric patterns and positional relationships within images. The SimAM module can learn latent vectors highlighting objects or local regions through an energy function using the activation values of neurons and surrounding neurons. The proposed method is a representation learning model that can generate low-dimensional latent vectors from high-dimensional input images, and the latent vectors are binarized into distinguishable hash code. From the experimental results on public datasets such as CIFAR-10, ImageNet, and NUS-WIDE, the proposed model is superior to the comparative model and analyzed to have equivalent performance to the supervised learning-based deephashing model. The proposed model can be used in application systems that require low-dimensional representation of images, such as image search or copyright image determination.

• Journal of fish pathology
• /
• v.36 no.2
• /
• pp.277-286
• /
• 2023

Korean rockfish, Sebastes schlegelii, is a representative bony fish that belongs to the family Scorpaenidae and the order Scorpaeniformes. It has high ecological and economic value and is widely cultivated in many East Asian countries, including South Korea, Japan and China. One of streptococci, Streptococcus iniae, is Gram-positive cocci with a negative reaction for catalase and oxidase. The Korean rockfish shows clinical signs when infected with S. iniae, such as body darkening, bleeding, enlarged kidneys, blurred eyes, abdominal distension, etc., ultimately leading to death. The Korean rockfish causes significant economic losses every year in South Korea due to streptococcosis. In this study, we identified bacteria from the fish using polymerase chain reaction and conducted analyses of hemolytic activity and biochemical tests using API 20 STREP and API ZYM systems. Results of confirming the hemolytic activity (n=4) observed in alpha-type hemolysis (25%), beta-type hemol- ysis (50%), and gamma-type hemolysis (25%) of isolates. The biochemical test results exhibited sig- nificant variation among S. iniae. Additionally, we performed intraperitoneal injection with S. iniae in the fish and analyzed the phylogenetic tree using housekeeping genes of S. iniae, including cpsD, arcC, glnA, groEL, gyrB, mutS, pheT, prkC, rpoB, and tkt, via multilocus sequence typing (MLST). The lethal dose (LD₅₀) showed strong pathogenicity, such as 3.34 × 10 colony-forming unit (CFU)/ml for 23FBStr0601 strain and 7.16 × 10 CFU/ml for 23FBStr0602 strain. 23FBStr0603 strain showed relatively low pathogenicity at 1.73 × 10⁵ CFU/ml. The strains 23FBStr0601 and 23FBStr0602, which showed strong pathogenicity, clustered into one monophyletic group. The 23FBStr0603 strain showed weak pathogenicity and formed a monophyletic group with KCTC 3657.

• Journal of Korea Water Resources Association
• /
• v.56 no.spc1
• /
• pp.1059-1069
• /
• 2023

Dam-break flow occurs when an elevated dam suddenly collapses, resulting in the catastrophic release of rapid and uncontrolled impounded water. This study compares laminar and turbulent closure models for simulating three-dimensional dam-break flows using OpenFOAM. The Reynolds-Averaged Navier-Stokes (RANS) model, specifically the k-ε model, is employed to capture turbulent dissipation. Two scenarios are evaluated based on a laboratory experiment and a modified multi-layered block obstacle scenario. Both models effectively represent dam-break flows, with the turbulent closure model reducing oscillations. However, excessive dissipation in turbulent models can underestimate water surface profiles. Improving numerical schemes and grid resolution enhances flow recreation, particularly near structures and during turbulence. Model stability is more significantly influenced by numerical schemes and grid refinement than the use of turbulence closure. The k-ε model's reliance on time-averaging processes poses challenges in representing dam-break profiles with pronounced discontinuities and unsteadiness. While simulating turbulence models requires extensive computational efforts, the performance improvement compared to laminar models is marginal. To achieve better representation, more advanced turbulence models like Large Eddy Simulation (LES) and Direct Numerical Simulation (DNS) are recommended, necessitating small spatial and time scales. This research provides insights into the applicability of different modeling approaches for simulating dam-break flows, emphasizing the importance of accurate representation near structures and during turbulence.

• The Journal of The Korea Institute of Intelligent Transport Systems
• /
• v.23 no.1
• /
• pp.153-166
• /
• 2024

The digitalization of tasks for smart construction requires the smooth exchange of digital data among stakeholders to be effective, but there is a lack of digital data standardization and utilization methods. This paper proposes a digital construction information structure to transform information from road construction sites into digital formats. The study targets include significant tasks, such as work planning, scheduling, safety management, and quality control. The key to the construction information structure is separating construction information into objects and activities, defining unit works by combining these two types of information to ensure flexibility in representing and modifying construction information. The objects and activities have their respective hierarchical structures, which are defined flexibly to match the actual content. This structure achieves both efficiency and detail. The pilot structure was applied to highway construction projects and implemented digitally using general formats. This study enables the digitalization of road construction processes that closely resemble reality, accelerating the digital transformation of the civil engineering industry by developing a digital twin of the entire road construction lifecycle.

• The Transactions of the Korea Information Processing Society
• /
• v.13 no.6
• /
• pp.284-290
• /
• 2024

Speech emotion recognition (SER) is a technique that is used to analyze the speaker's voice patterns, including vibration, intensity, and tone, to determine their emotional state. There has been an increase in interest in artificial intelligence (AI) techniques, which are now widely used in medicine, education, industry, and the military. Nevertheless, existing researchers have attained impressive results by utilizing acted-out speech from skilled actors in a controlled environment for various scenarios. In particular, there is a mismatch between acted and spontaneous speech since acted speech includes more explicit emotional expressions than spontaneous speech. For this reason, spontaneous speech-emotion recognition remains a challenging task. This paper aims to conduct emotion recognition and improve performance using spontaneous speech data. To this end, we implement deep learning-based speech emotion recognition using the VGG (Visual Geometry Group) after converting 1-dimensional audio signals into a 2-dimensional spectrogram image. The experimental evaluations are performed on the Korean spontaneous emotional speech database from AI-Hub, consisting of 7 emotions, i.e., joy, love, anger, fear, sadness, surprise, and neutral. As a result, we achieved an average accuracy of 83.5% and 73.0% for adults and young people using a time-frequency 2-dimension spectrogram, respectively. In conclusion, our findings demonstrated that the suggested framework outperformed current state-of-the-art techniques for spontaneous speech and showed a promising performance despite the difficulty in quantifying spontaneous speech emotional expression.

• Journal of The Korean Association For Science Education
• /
• v.26 no.4
• /
• pp.475-491
• /
• 2006

In physics education, the research on students' conceptions has developed in the discussion on the nature and the difficulty of conceptual change. Recently, mental models have been a theoretical background in concrete arguments on "how students' conceptions are constructed or created." Mental models that integrate information in the presented problem and individual knowledge in their long-term memory have important information about not only expressed ideas but also in the thinking process behind the expressed ideas. The purpose of this study is to investigate the forming process and the characteristics of high school student's mental models about circular motion, and how they were changed by instruction. We used the think-aloud method based on the instrument for identifying student's mental models about circular motion, pretest of physics concept, mind map and interview for investigating student's characteristics. The results of the study showed that instructions based on the mental model theory facilitated scientific expressed model, but several factors that affected forming mental models like epistemological belief didn't change scientifically after 3 lessons.

• Journal of The Korean Association For Science Education
• /
• v.29 no.8
• /
• pp.978-989
• /
• 2009

In this study, we investigated high school students' errors in constructing and interpreting graph on experimental results by students' science achievement level. Two tests regarding constructing and interpreting graph about 'the relationship between the pressure and volume of a gas' were administered to 11th-graders (N=140). Analysis of the results revealed that most students exhibited many errors in the processes of constructing and interpreting graph. In the processes of constructing graph, there were 16 types of errors on the categories of 'misinterpreting the variables', 'mis-marking the graphical elements', and 'misusing the data'. The students of lower achievement level had more errors than those of higher achievement level in the four error types, that is, 'missing the variables', 'representing the best fit line using a broken line', 'adding the data', and 'neglecting the data'. However, the results were reversed in the error type of 'not marking the origin.' In the processes of interpreting graph, there were 9 types of errors on 'misreading the data', 'wrong interpolation and extrapolation', and 'establishing the wrong relationship'. The students of lower achievement level had more errors than those in the higher achievement level in the error types of 'wrong interpolation' and 'misdescribing the relationship between variables'. Educational implications of the findings are discussed.

• The Sea:JOURNAL OF THE KOREAN SOCIETY OF OCEANOGRAPHY
• /
• v.15 no.2
• /
• pp.86-96
• /
• 2010

The oxygen utilization rate (OUR) is one of the crucial parameters for ocean carbon cycling and climate models. However, parameterization of OUR in the East Sea (Sea of Japan) is yet to be established. We estimated the basin-specific OURs in the East Sea and fitted them with exponential functions with depth by using pCFC- 12 age and apparent oxygen utilization (AOU) measured in summer 1999. The estimated OURs are higher in the upper water column and decrease with depth, in general. The vertical distributions of the estimated OURs in the Western and Eastern Japan Basins (WJB & EJB) are very similar. The OURs in the Ulleung Basin (UB) varied greatly depending on whether the surface layer (0~200 m) data are included in the OUR estimate or not. Apparently, weaker oxygen consumption occurs in the deep layer of Yamato Basin (YB). The ranges of the OURs between 200 m and 2000 m at WJB, EJB, UB, and YB are 8.15~0.83, 8.11~0.68, 5.29~0.73, and 7.31~0.06 ${\mu}mol$ $kg^{-1}$ $yr^{-1}$, respectively. Consideration of the wintertime surface water oxygen disequilibrium condition in estimating the OUR will be necessary in the future study.

• Journal of The Korean Association For Science Education
• /
• v.28 no.1
• /
• pp.32-46
• /
• 2008

The purpose of this study was to analyse the qualitative and quantitative characteristics of paper-pencil tests for exploring its appropriateness as a selection tool of the gifted in science. For this purpose, we developed two (internal and external) item analysis frameworks, and applied these frameworks to analyse qualitative characteristics. Also, we analysed the relationship between two characteristics. The results of analysing qualitative characteristics revealed that the portion of items with acceleration context exceeding middle school curriculum level was relatively large, which caused low content validity. Furthermore, there was considerable deviation in content and context by subject matter and year, which caused test unstability. Items measuring knowledge domain was the most prevalent, and too much weight on data interpretation & analysis domain in inquiry process skills. In case of creativity test, the portion of items measuring convergent thinking was much larger than that of divergent or associative thinking. Most of these items were represented by using pictures and tables rather than using graphs. Item types of multiple-choice and short answers were superior to essay types. Discrimination index, on the whole, was appropriate (above 0.3), but item difficulty showed a vast deviation ($0.01{\sim}0.90$). Correlation coefficients among subject matters and test tools were very low, and test reliabilities were also low. Low item difficulty & high discrimination index item types were distinguishable. Items with acceleration context were more discriminating than enrichment context. Implications of developing quality paper-pencil test items in the selection of gifted students are discussed.

• Journal of The Korean Association For Science Education
• /
• v.27 no.9
• /
• pp.809-817
• /
• 2007

The purpose of this research was to investigate the nature of the science views of science teachers and high school students as well as the views expressed in 10th grade science textbooks. The subjects were a high school science teacher, 18 male students and 11 female students in a 10th grade class located in Gyeongbuk Province, Korea. The data were analyzed in terms of three main categories of the nature of the science: the definition of science, the development of science, and the method of science. In the results, it was found that the textbook had an inductivism point of view, and that the teacher had a falsificationism point of view in terms of the definition of science. However, the teacher presented the inductive point of view of the textbook in the class. After the class, the students showed an inductive point of view. In terms of scientific development, the textbook represented a relativism point of view briefly, and the viewpoint of the teacher also expressed relativism. The teacher taught briefly from the relativism point of view, as in the textbook. The viewpoints of the students were inductivism and were not affected by the textbook or the explanations of the teacher. In terms of scientific methods, the viewpoints of the textbook and the teacher were falsificationism, and the teacher represented falsificationism views in her classes. The views of the students also showed falsificationism after their classes. However, before conclusions can be made, it is necessary to find concrete proof of the teaching effect on the viewpoints of the students in continuing research.