• Korean Journal of Remote Sensing
• /
• v.40 no.4
• /
• pp.343-350
• /
• 2024

Since the release of Meta's Segment Anything Model (SAM), a large-scale vision transformer generation model with rapid image segmentation capabilities, several studies have been conducted to apply this technology in various fields. In this study, we aimed to investigate the applicability of SAM for water bodies detection and extraction using the QGIS Geo-SAM plugin, which enables the use of SAM with satellite imagery. The experimental data consisted of Compact Advanced Satellite 500 (CAS500)-1 images. The results obtained by applying SAM to these data were compared with manually digitized water objects, Open Street Map (OSM), and water body data from the National Geographic Information Institute (NGII)-based hydrological digital map. The mean Intersection over Union (mIoU) calculated for all features extracted using SAM and these three-comparison data were 0.7490, 0.5905, and 0.4921, respectively. For features commonly appeared or extracted in all datasets, the results were 0.9189, 0.8779, and 0.7715, respectively. Based on analysis of the spatial consistency between SAM results and other comparison data, SAM showed limitations in detecting small-scale or poorly defined streams but provided meaningful segmentation results for water body classification.

• Journal of the Korea Society of Computer and Information
• /
• v.29 no.8
• /
• pp.11-21
• /
• 2024

This study explores the use of virtual reality (VR) technology to identify and label key segments of the golf swing. To address the limitations of existing VR devices, we developed a platform to collect kinematic data from various VR devices using the OpenVR SDK (Software Development Kit) and SteamVR, and developed a semi-automated labeling technique to identify and label temporal changes in kinematic behavior through LSTM (Long Short-Term Memory)-based time series data analysis. The experiment consisted of 80 participants, 20 from each of the following age groups: teenage, young-adult, middle-aged, and elderly, collecting data from five swings each to build a total of 400 kinematic datasets. The proposed technique achieved consistently high accuracy (≥0.94) and F1 Score (≥0.95) across all age groups for the seven main phases of the golf swing. This work aims to lay the groundwork for segmenting exercise data and precisely assessing athletic performance on a segment-by-segment basis, thereby providing personalized feedback to individual users during future education and training.

• Journal of the Korea Society of Computer and Information
• /
• v.29 no.8
• /
• pp.157-164
• /
• 2024

In this paper, we propose a discussion that the feasibility of deploying a deep learning-based detector on the resource-limited board. Although many studies evaluate the detector on machines with high-performed GPUs, evaluation on the board with limited computation resources is still insufficient. Therefore, in this work, we implement the deep-learning detectors and deploy them on the compact board by parsing and optimizing a detector. To figure out the performance of deep learning based detectors on limited resources, we monitor the performance of several detectors with different H/W resource. On COCO detection datasets, we compare and analyze the evaluation results of detection model in On-Board and the detection model in On-GPU in terms of several metrics with mAP, power consumption, and execution speed (FPS). To demonstrate the effect of applying our detector for the military area, we evaluate them on our dataset consisting of thermal images considering the flight battle scenarios. As a results, we investigate the strength of deep learning-based on-board detector, and show that deep learning-based vision models can contribute in the flight battle scenarios.

• Journal of Wetlands Research
• /
• v.26 no.3
• /
• pp.197-203
• /
• 2024

The algal blooms in rivers can negatively affect water source management and water treatment processes, necessitating continuous management. In this study, a multi-classification model was developed to predict the concentration of chlorophyll-a (chl-a), one of the key indicators of algal blooms, using Tabular Prior Fitted Networks (TabPFN), a novel deep learning algorithm known for its relatively superior performance on small tabular datasets. The model was developed using daily observation data collected at Buyeo water quality monitoring station from January 1, 2014, to December 31, 2022. The collected data were averaged to construct input data sets with measurement frequencies of 1 day, 3 days, 6 days, 12 days. The performance comparison of the four models, constructed with input data on observation frequencies of 1 day, 3 days, 6 days, and 12 days, showed that the model exhibits stable performance even when the measurement frequency is longer and the number of observations is smaller. The macro average for each model were analyzed as follows: Precision was 0.77, 0.76, 0.83, 0.84; Recall was 0.63, 0.65, 0.66, 0.74; F1-score was 0.67, 0.69, 0.71, 0.78. For the weighted average, Precision was 0.76, 0.77, 0.81, 0.84; Recall was 0.76, 0.78, 0.81, 0.85; F1-score was 0.74, 0.77, 0.80, 0.84. This study demonstrates that the chl-a prediction model constructed using TabPFN exhibits stable performance even with small-scale input data, verifying the feasibility of its application in fields where the input data required for model construction is limited.

• Journal of Intelligence and Information Systems
• /
• v.24 no.1
• /
• pp.205-225
• /
• 2018

Convolutional Neural Network (ConvNet) is one class of the powerful Deep Neural Network that can analyze and learn hierarchies of visual features. Originally, first neural network (Neocognitron) was introduced in the 80s. At that time, the neural network was not broadly used in both industry and academic field by cause of large-scale dataset shortage and low computational power. However, after a few decades later in 2012, Krizhevsky made a breakthrough on ILSVRC-12 visual recognition competition using Convolutional Neural Network. That breakthrough revived people interest in the neural network. The success of Convolutional Neural Network is achieved with two main factors. First of them is the emergence of advanced hardware (GPUs) for sufficient parallel computation. Second is the availability of large-scale datasets such as ImageNet (ILSVRC) dataset for training. Unfortunately, many new domains are bottlenecked by these factors. For most domains, it is difficult and requires lots of effort to gather large-scale dataset to train a ConvNet. Moreover, even if we have a large-scale dataset, training ConvNet from scratch is required expensive resource and time-consuming. These two obstacles can be solved by using transfer learning. Transfer learning is a method for transferring the knowledge from a source domain to new domain. There are two major Transfer learning cases. First one is ConvNet as fixed feature extractor, and the second one is Fine-tune the ConvNet on a new dataset. In the first case, using pre-trained ConvNet (such as on ImageNet) to compute feed-forward activations of the image into the ConvNet and extract activation features from specific layers. In the second case, replacing and retraining the ConvNet classifier on the new dataset, then fine-tune the weights of the pre-trained network with the backpropagation. In this paper, we focus on using multiple ConvNet layers as a fixed feature extractor only. However, applying features with high dimensional complexity that is directly extracted from multiple ConvNet layers is still a challenging problem. We observe that features extracted from multiple ConvNet layers address the different characteristics of the image which means better representation could be obtained by finding the optimal combination of multiple ConvNet layers. Based on that observation, we propose to employ multiple ConvNet layer representations for transfer learning instead of a single ConvNet layer representation. Overall, our primary pipeline has three steps. Firstly, images from target task are given as input to ConvNet, then that image will be feed-forwarded into pre-trained AlexNet, and the activation features from three fully connected convolutional layers are extracted. Secondly, activation features of three ConvNet layers are concatenated to obtain multiple ConvNet layers representation because it will gain more information about an image. When three fully connected layer features concatenated, the occurring image representation would have 9192 (4096+4096+1000) dimension features. However, features extracted from multiple ConvNet layers are redundant and noisy since they are extracted from the same ConvNet. Thus, a third step, we will use Principal Component Analysis (PCA) to select salient features before the training phase. When salient features are obtained, the classifier can classify image more accurately, and the performance of transfer learning can be improved. To evaluate proposed method, experiments are conducted in three standard datasets (Caltech-256, VOC07, and SUN397) to compare multiple ConvNet layer representations against single ConvNet layer representation by using PCA for feature selection and dimension reduction. Our experiments demonstrated the importance of feature selection for multiple ConvNet layer representation. Moreover, our proposed approach achieved 75.6% accuracy compared to 73.9% accuracy achieved by FC7 layer on the Caltech-256 dataset, 73.1% accuracy compared to 69.2% accuracy achieved by FC8 layer on the VOC07 dataset, 52.2% accuracy compared to 48.7% accuracy achieved by FC7 layer on the SUN397 dataset. We also showed that our proposed approach achieved superior performance, 2.8%, 2.1% and 3.1% accuracy improvement on Caltech-256, VOC07, and SUN397 dataset respectively compare to existing work.

• Journal of Internet Computing and Services
• /
• v.21 no.4
• /
• pp.17-23
• /
• 2020

Biometric information indicating measurement items related to human characteristics has attracted great attention as security technology with high reliability since there is no fear of theft or loss. Among these biometric information, fingerprints are mainly used in fields such as identity verification and identification. If there is a problem such as a wound, wrinkle, or moisture that is difficult to authenticate to the fingerprint image when identifying the identity, the fingerprint expert can identify the problem with the fingerprint directly through the preprocessing step, and apply the image processing algorithm appropriate to the problem. Solve the problem. In this case, by implementing artificial intelligence software that distinguishes fingerprint images with cuts and wrinkles on the fingerprint, it is easy to check whether there are cuts or wrinkles, and by selecting an appropriate algorithm, the fingerprint image can be easily improved. In this study, we developed a total of 17,080 fingerprint databases by acquiring all finger prints of 1,010 students from the Royal University of Cambodia, 600 Sokoto open data sets, and 98 Korean students. In order to determine if there are any injuries or wrinkles in the built database, criteria were established, and the data were validated by experts. The training and test datasets consisted of Cambodian data and Sokoto data, and the ratio was set to 8: 2. The data of 98 Korean students were set up as a validation data set. Using the constructed data set, five CNN-based architectures such as Classic CNN, AlexNet, VGG-16, Resnet50, and Yolo v3 were implemented. A study was conducted to find the model that performed best on the readings. Among the five architectures, ResNet50 showed the best performance with 81.51%.

• Journal of Internet Computing and Services
• /
• v.16 no.6
• /
• pp.89-95
• /
• 2015

Traditional frequent pattern mining discovers valid patterns with no smaller frequency than a user-defined minimum threshold from databases. In this framework, an enormous number of patterns may be extracted by a too low threshold, which makes result analysis difficult, and a too high one may generate no valid pattern. Setting an appropriate threshold is not an easy task since it requires the prior knowledge for its domain. Therefore, a pattern mining approach that is not based on the domain knowledge became needed due to inability of the framework to predict and control mining results precisely according to the given threshold. Top-k frequent pattern mining was proposed to solve the problem, and it mines top-k important patterns without any threshold setting. Through this method, users can find patterns from ones with the highest frequency to ones with the k-th highest frequency regardless of databases. In this paper, we provide knowledge both on frequent and top-k pattern mining. Although top-k frequent pattern mining extracts top-k significant patterns without the setting, it cannot consider both item quantities in transactions and relative importance of items in databases, and this is why the method cannot meet requirements of many real-world applications. That is, patterns with low frequency can be meaningful, and vice versa, in the applications. High utility pattern mining was proposed to reflect the characteristics of non-binary databases and requires a minimum threshold. Recently, top-k high utility pattern mining has been developed, through which users can mine the desired number of high utility patterns without the prior knowledge. In this paper, we analyze two algorithms related to top-k high utility pattern mining in detail. We also conduct various experiments for the algorithms on real datasets and study improvement point and development direction of top-k high utility pattern mining through performance analysis with respect to the experimental results.

• Journal of Intelligence and Information Systems
• /
• v.24 no.3
• /
• pp.21-44
• /
• 2018

In recent years, the rapid development of internet technology and the popularization of smart devices have resulted in massive amounts of text data. Those text data were produced and distributed through various media platforms such as World Wide Web, Internet news feeds, microblog, and social media. However, this enormous amount of easily obtained information is lack of organization. Therefore, this problem has raised the interest of many researchers in order to manage this huge amount of information. Further, this problem also required professionals that are capable of classifying relevant information and hence text classification is introduced. Text classification is a challenging task in modern data analysis, which it needs to assign a text document into one or more predefined categories or classes. In text classification field, there are different kinds of techniques available such as K-Nearest Neighbor, Naïve Bayes Algorithm, Support Vector Machine, Decision Tree, and Artificial Neural Network. However, while dealing with huge amount of text data, model performance and accuracy becomes a challenge. According to the type of words used in the corpus and type of features created for classification, the performance of a text classification model can be varied. Most of the attempts are been made based on proposing a new algorithm or modifying an existing algorithm. This kind of research can be said already reached their certain limitations for further improvements. In this study, aside from proposing a new algorithm or modifying the algorithm, we focus on searching a way to modify the use of data. It is widely known that classifier performance is influenced by the quality of training data upon which this classifier is built. The real world datasets in most of the time contain noise, or in other words noisy data, these can actually affect the decision made by the classifiers built from these data. In this study, we consider that the data from different domains, which is heterogeneous data might have the characteristics of noise which can be utilized in the classification process. In order to build the classifier, machine learning algorithm is performed based on the assumption that the characteristics of training data and target data are the same or very similar to each other. However, in the case of unstructured data such as text, the features are determined according to the vocabularies included in the document. If the viewpoints of the learning data and target data are different, the features may be appearing different between these two data. In this study, we attempt to improve the classification accuracy by strengthening the robustness of the document classifier through artificially injecting the noise into the process of constructing the document classifier. With data coming from various kind of sources, these data are likely formatted differently. These cause difficulties for traditional machine learning algorithms because they are not developed to recognize different type of data representation at one time and to put them together in same generalization. Therefore, in order to utilize heterogeneous data in the learning process of document classifier, we apply semi-supervised learning in our study. However, unlabeled data might have the possibility to degrade the performance of the document classifier. Therefore, we further proposed a method called Rule Selection-Based Ensemble Semi-Supervised Learning Algorithm (RSESLA) to select only the documents that contributing to the accuracy improvement of the classifier. RSESLA creates multiple views by manipulating the features using different types of classification models and different types of heterogeneous data. The most confident classification rules will be selected and applied for the final decision making. In this paper, three different types of real-world data sources were used, which are news, twitter and blogs.

• Journal of Korea Water Resources Association
• /
• v.47 no.4
• /
• pp.371-384
• /
• 2014

This study developed a new algorithm of extreme rainfall extraction based on the Communication, Ocean and Meteorological Satellite (COMS) and the Tropical Rainfall Measurement Mission (TRMM) Satellite image data and evaluated its applicability for the heavy rainfall event in July-2011 in Seoul, South Korea. The power-series-regression-based Z-R relationship was employed for taking into account for empirical relationships between TRMM/PR, TRMM/VIRS, COMS, and Automatic Weather System(AWS) at each elevation. The estimated Z-R relationship ($Z=303R^{0.72}$) agreed well with observation from AWS (correlation coefficient=0.57). The estimated 10-minute rainfall intensities from the COMS satellite using the Z-R relationship generated underestimated rainfall intensities. For a small rainfall event the Z-R relationship tended to overestimated rainfall intensities. However, the overall patterns of estimated rainfall were very comparable with the observed data. The correlation coefficients and the Root Mean Square Error (RMSE) of 10-minute rainfall series from COMS and AWS gave 0.517, and 3.146, respectively. In addition, the averaged error value of the spatial correlation matrix ranged from -0.530 to -0.228, indicating negative correlation. To reduce the error by extreme rainfall estimation using satellite datasets it is required to take into more extreme factors and improve the algorithm through further study. This study showed the potential utility of multi-geostationary satellite data for building up sub-daily rainfall and establishing the real-time flood alert system in ungauged watersheds.

• Journal of the Korean Association of Geographic Information Studies
• /
• v.10 no.3
• /
• pp.80-92
• /
• 2007

Forest fire is the dominant large-scale disturbance mechanism in the Korean temperate forest, and it strongly influences forest structure and function. Moreover burn severity incorporates both short- and long-term post-fire effects on the local and regional environment. Burn severity is defined by the degree to which an ecosystem has changed owing to the fire. Vegetation rehabilitation may specifically vary according to burn severity after fire. To understand burn severity and process of vegetation rehabilitation at the damaged area after large-fire is required a lot of man powers and budgets. However the analysis of burn severity in the forest area using satellite imagery can acquire rapidly information and more objective results remotely in the large-fire area. Space and airbone sensors have been used to map area burned, assess characteristics of active fires, and characterize post-fire ecological effects. For classifying fire damaged area and analyzing burn severity of Samcheok fire area occurred in 2000, Cheongyang fire in 2002, and Yangyang fire in 2005 we utilized Normalized Burn Ratio(NBR) technique. The NBR is temporally differenced between pre- and post-fire datasets to determine the extent and degree of change detected from burning. In this paper we use pre- and post-fire imagery from the Landsat TM and ETM+ imagery to compute the NBR and evaluate large-scale patterns of burn severity at 30m spatial resolution. 65% in the Samcheok fire area, 91% in the Cheongyang fire area and 65% in the Yangyang fire area were corresponded to burn severity class above 'High'. Therefore the use of a remotely sensed Differenced Normalized Burn Ratio(${\Delta}NBR$) by RS and GIS allows for the burn severity to be quantified spatially by mapping damaged domain and burn severity across large-fire area.