• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.9 no.2
• /
• pp.775-792
• /
• 2015

Person re-identification is an important and challenging task in computer vision with numerous real world applications. Despite significant progress has been made in the past few years, person re-identification remains an unsolved problem. This paper presents a novel appearance-based approach to person re-identification. The approach exploits region covariance matrix and color histograms to capture the statistical properties and chromatic information of each object. Robustness against low resolution, viewpoint changes and pose variations is achieved by a novel signature, that is, the combination of Log Covariance Matrix feature and HSV histogram (LCMH). In order to further improve re-identification performance, third-party image sets are utilized as a common reference to sufficiently represent any image set with the same type. Distinctive and reliable features for a given image set are extracted through decision boundary between the specific set and a third-party image set supervised by max-margin criteria. This method enables the usage of an existing dataset to represent new image data without time-consuming data collection and annotation. Comparisons with state-of-the-art methods carried out on benchmark datasets demonstrate promising performance of our method.

• Journal of the Korean Society for information Management
• /
• v.33 no.1
• /
• pp.317-336
• /
• 2016

Nowadays users spend more time and effort to find what they want because of information overload. To solve the problem, scientific article recommendation system analyse users' needs and recommend them proper articles. However, most of the scientific article recommendation systems neglected the core part, user profile. Therefore, in this paper, instead of mean which applied in user profile in previous studies, New TPIPF (Topic Proportion-Inverse Paper Frequency) was applied to scientific article recommendation system. Moreover, the accuracy of two scientific article recommendation systems with above different methods was compared with experiments of public dataset from online reference manager, CiteULike. As a result, the proposed scientific article recommendation system with TPIPF was proven to be better.

• Journal of the Korean Society for information Management
• /
• v.27 no.1
• /
• pp.103-118
• /
• 2010

Recently, the valuation of patented technology has been greatly emphasized, and patent citation has been accepted as a very useful index of this technology. In this study, we performed correlation analyses between the patent citation counts and 17 explanatory variables of morphological, technological, and conceptual factors with a test dataset of U.S. patents in five subject fields. Seven variables having 5% or more standardized variances($r^2$) with patent citation counts were identified; number of pages, number of claims, reference-average-citation rate, patent increase/decrease rate, strength of bibliographic coupling, co-citation counts and document similarity. The result of the ANOVA test shows that the mean values of these variables vary among most subject fields.

• Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography
• /
• v.36 no.1
• /
• pp.17-24
• /
• 2018

DEM (Digital Elevation Model) is a useful dataset which represents the earth surface. Beside many applications, production and frequent update of DEM is a costly task. Recently global satellite based DEMs are available which has huge potential for application. To check the accuracy, this study compares two global DEMs: AW3D30 (Advanced Land Observing Satellite World 3D 30m) and SRTM30 (Shuttle Radar Topography Mission Global 30m) with reference resampled LiDAR DEM 30m in a test area around Chuncheon, Korea. The comparison analysis was based on statistics of each DEM, their difference, profiles, slope, basin and stream orders. As a result, it is found that SRTM30 and AW3D30 were much similar but inconsistent in the test area compared to the LiDAR30 DEM. In addition, SRTM30 shows less difference with LiDAR30 compared to the AW3D30 DEM. But, DEMs should be very carefully examined for area which has temporal or season changes. For basin and stream analysis, global DEMs can be used only for regional scale analysis not local large scales.

• Journal of Astronomy and Space Sciences
• /
• v.34 no.1
• /
• pp.7-17
• /
• 2017

In South Korea, there are about 80 Global Positioning System (GPS) monitoring stations providing total electron content (TEC) every 10 min, which can be accessed through Korea Astronomy and Space Science Institute (KASI) for scientific use. We applied the computerized ionospheric tomography (CIT) algorithm to the TEC dataset from this GPS network for monitoring the regional ionosphere over South Korea. The algorithm utilizes multiplicative algebraic reconstruction technique (MART) with an initial condition of the latest International Reference Ionosphere-2016 model (IRI-2016). In order to reduce the number of unknown variables, the vertical profiles of electron density are expressed with a linear combination of empirical orthonormal functions (EOFs) that were derived from the IRI empirical profiles. Although the number of receiver sites is much smaller than that of Japan, the CIT algorithm yielded reasonable structure of the ionosphere over South Korea. We verified the CIT results with NmF2 from ionosondes in Icheon and Jeju and also with GPS TEC at the center of South Korea. In addition, the total time required for CIT calculation was only about 5 min, enabling the exploration of the vertical ionospheric structure in near real time.

• Korean Journal of Fisheries and Aquatic Sciences
• /
• v.53 no.5
• /
• pp.740-751
• /
• 2020

In the aquaculture industry, few studies are analyzing big data for intrinsic meaning. Fishcare Laboratory (www.fishcare.kr) diagnostic data from 2016-2018 was analyzed for scuticociliatosis (caused by Miamiensis avidus) outbreak patterns in cultured olive flounder Paralichthys olivaceus in Jeju, Korea. The scuticociliatosis monthly occurrence ratio is reported in the summary table after preparing and filtering the basic dataset model. Nonparametric test results suggest differences in the water temperature, body length, and weight between groups with and without scuticociliatosis. Data distribution visualization revealed that shorter body length and lighter weight increased the occurrence of scuticociliatosis. The association rule mining technique was applied to determine the primary clinical signs of mixed scuticociliatosis and bacterial infections. Venn diagrams were used to report clinical signs and suggest commonalities. These results may help diagnose and treat fish and provide a decision-making reference.

• The Transactions of The Korean Institute of Electrical Engineers
• /
• v.63 no.6
• /
• pp.797-803
• /
• 2014

In this paper, 3D face recognition system is designed by using polynomial based on RBFNNs. In case of 2D face recognition, the recognition performance reduced by the external environmental factors such as illumination and facial pose. In order to compensate for these shortcomings of 2D face recognition, 3D face recognition. In the preprocessing part, according to the change of each position angle the obtained 3D face image shapes are changed into front image shapes through pose compensation. the depth data of face image shape by using Multiple Point Signature is extracted. Overall face depth information is obtained by using two or more reference points. The direct use of the extracted data an high-dimensional data leads to the deterioration of learning speed as well as recognition performance. We exploit principle component analysis(PCA) algorithm to conduct the dimension reduction of high-dimensional data. Parameter optimization is carried out with the aid of PSO for effective training and recognition. The proposed pattern classifier is experimented with and evaluated by using dataset obtained in IC & CI Lab.

• Journal of Microbiology and Biotechnology
• /
• v.26 no.6
• /
• pp.1124-1131
• /
• 2016

In this study, the global proteome of the IPEC-J2 cell line was evaluated using ultra-high performance liquid chromatography coupled to a quadrupole Q Exactive^™ Orbitrap mass spectrometer. Proteins were isolated from highly confluent IPEC-J2 cells in biological replicates and analyzed by label-free mass spectrometry prior to matching against a porcine genomic dataset. The results identified 1,517 proteins, accounting for 7.35% of all genes in the porcine genome. The highly abundant proteins detected, such as actin, annexin A2, and AHNAK nucleoprotein, are involved in structural integrity, signaling mechanisms, and cellular homeostasis. The high abundance of heat shock proteins indicated their significance in cellular defenses, barrier function, and gut homeostasis. Pathway analysis and annotation using the Kyoto Encyclopedia of Genes and Genomes database resulted in a putative protein network map of the regulation of immunological responses and structural integrity in the cell line. The comprehensive proteome analysis of IPEC-J2 cells provides fundamental insights into overall protein expression and pathway dynamics that might be useful in cell adhesion studies and immunological applications.

• Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography
• /
• v.37 no.5
• /
• pp.331-341
• /
• 2019

With the popularity of sensor-rich environments, smartphones have become one of the major platforms for obtaining and sharing information. Since it is difficult to utilize GNSS (Global Navigation Satellite System) inside the area with many buildings, the localization of smartphone in this case is considered as a challenging task. To resolve problem of localization using smartphone a four step image-based localization method and procedure is proposed. To improve the localization accuracy of smartphone datasets, MMS (Mobile Mapping System) and Google Street View were utilized. In our approach first, the searching for candidate matching image is performed by the query image of smartphone's using GNSS observation. Second, the SURF (Speed-Up Robust Features) image matching between the smartphone image and reference dataset is done and the wrong matching points are eliminated. Third, the geometric transformation is performed using the matching points with 2D affine transformation. Finally, the smartphone location and attitude estimation are done by PnP (Perspective-n-Point) algorithm. The location of smartphone GNSS observation is improved from the original 10.204m to a mean error of 3.575m. The attitude estimation is lower than 25 degrees from the 92.4% of the adjsuted images with an average of 5.1973 degrees.

• Journal of Information Processing Systems
• /
• v.17 no.5
• /
• pp.867-878
• /
• 2021

In context awareness and user intention tasks, dataset construction is expensive because specific domain data are required. Although pretraining with a large corpus can effectively resolve the issue of lack of data, it ignores domain knowledge. Herein, we concentrate on data domain knowledge while addressing data scarcity and accordingly propose a multi-channel long short-term memory (LSTM). Because multi-channel LSTM integrates pretrained vectors such as task and general knowledge, it effectively prevents catastrophic forgetting between vectors of task and general knowledge to represent the context as a set of features. To evaluate the proposed model with reference to the baseline model, which is a single-channel LSTM, we performed two tasks: voice phishing with context awareness and movie review sentiment classification. The results verified that multi-channel LSTM outperforms single-channel LSTM in both tasks. We further experimented on different multi-channel LSTMs depending on the domain and data size of general knowledge in the model and confirmed that the effect of multi-channel LSTM integrating the two types of knowledge from downstream task data and raw data to overcome the lack of data.