• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.30 no.12C
• /
• pp.1162-1166
• /
• 2005

A new low dimensional feature representation technique is presented in this paper. Linear discriminant analysis is a popular feature extraction method. However, in the case of high dimensional data, the computational difficulty and the small sample size problem are often encountered. In order to solve these problems, we propose two dimensional direct LDA algorithm, which directly extracts the image scatter matrix from 2D image and uses Direct LDA algorithm for face recognition. The ORL face database is used to evaluate the performance of the proposed method. The experimental results indicate that the performance of the proposed method is superior to DLDA.

• The Journal of the Acoustical Society of Korea
• /
• v.16 no.1
• /
• pp.100-103
• /
• 1997

In this paper, we analyze the time delayed effect of speech feature. Here, the time delayed effect means that the current feature vector of speech is under the influence of the previous feature vectors. In this paper, we use a set of LPC driven cepstal coefficients and evaluate the time delayed effect of cepstrum with the performance of the speech recognition system. For the experiments, we used the speech database consisting of 22 words which uttered by 50 male speakers. The speech database uttered by 25 male speakers was used for training, and the other set was used for testing. The experimental results show that the time delayed effect is large in the lower orders of feature vector but small in the higher orders.

• Journal of IKEEE
• /
• v.25 no.2
• /
• pp.337-343
• /
• 2021

Deep reinforcement learning is an artificial intelligence algorithm that enables learners to select optimal behavior based on raw and, high-dimensional input data. A lot of research using this is being conducted to create an optimal movement path of a mobile robot in an environment in which obstacles exist. In this paper, we selected the Dueling Double DQN (D3QN) algorithm that uses the prioritized experience replay to create the moving path of mobile robot from the image of the complex surrounding environment. The virtual environment is implemented using Webots, a robot simulator, and through simulation, it is confirmed that the mobile robot grasped the position of the obstacle in real time and avoided it to reach the destination.

• Journal of Korea Multimedia Society
• /
• v.25 no.6
• /
• pp.886-893
• /
• 2022

Outlier detection refers to the task of detecting data that deviate significantly from the normal data distribution. Most outlier detection methods compute an outlier score which indicates the degree to which a data sample deviates from normal. However, setting a threshold for an outlier score to determine if a data sample is outlier or normal is not trivial. In this paper, we propose a binary prediction method for outlier detection based on spectral clustering and one-class SVM ensemble. Given training data consisting of normal data samples, a clustering method is performed to find clusters in the training data, and the ensemble of one-class SVM models trained on each cluster finds the boundaries of the normal data. We show how to obtain a threshold for transforming outlier scores computed from the ensemble of one-class SVM models into binary predictive values. Experimental results with high dimensional text data show that the proposed method can be effectively applied to high dimensional data, especially when the normal training data consists of different shapes and densities of clusters.

• Journal of Korea Society of Industrial Information Systems
• /
• v.15 no.2
• /
• pp.49-57
• /
• 2010

This paper proposes a Korean document. categorization algorithm using Back Propagation Neural Network(BPNN) with Singular Value Decomposition(SVD). BPNN makes a network through its learning process and classifies documents using the network. The main difficulty in the application of BPNN to document categorization is high dimensionality of the feature space of the input documents. SVD projects the original high dimensional vector into low dimensional vector, makes the important associative relationship between terms and constructs the semantic vector space. The categorization algorithm is tested and compared on HKIB-20000/HKIB-40075 Korean Text Categorization Test Collections. Experimental results show that BPNN algorithm with SVD achieves high effectiveness for Korean document categorization.

• KIPS Transactions on Software and Data Engineering
• /
• v.8 no.10
• /
• pp.397-402
• /
• 2019

Accurately predicting cancer prognosis to provide appropriate treatment strategies for patients is one of the critical challenges in bioinformatics. Many researches have suggested machine learning models to predict patients' outcomes based on their gene expression data. Gene expression data is high-dimensional numerical data containing about 17,000 genes, so traditional researches used feature selection or dimensionality reduction approaches to elevate the performance of prognostic prediction models. These approaches, however, have an issue of making it difficult for the predictive models to grasp any biological interaction between the selected genes because feature selection and model training stages are performed independently. In this paper, we propose a novel two-dimensional image formatting approach for gene expression data to achieve feature selection and prognostic prediction effectively. Node2Vec is exploited to integrate biological interaction network and gene expression data and a convolutional neural network learns the integrated two-dimensional gene expression image data and predicts cancer prognosis. We evaluated our proposed model through double cross-validation and confirmed superior prognostic prediction accuracy to traditional machine learning models based on raw gene expression data. As our proposed approach is able to improve prediction models without loss of information caused by feature selection steps, we expect this will contribute to development of personalized medicine.

• Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography
• /
• v.40 no.2
• /
• pp.91-108
• /
• 2022

In most cases, optical images have been used as training data of DL (Deep Learning) models for object detection, recognition, identification, classification, semantic segmentation, and instance segmentation. However, properties of 3D objects in the real-world could not be fully explored with 2D images. One of the major sources of the 3D geospatial information is DSM (Digital Surface Model). In this matter, characteristic information derived from DSM would be effective to analyze 3D terrain features. Especially, man-made objects such as buildings having geometrically unique shape could be described by geometric elements that are obtained from 3D geospatial data. The background and motivation of this paper were drawn from concept of the intrinsic image that is involved in high-level visual information processing. This paper aims to extract buildings after classifying terrain features by training DL model with DSM-derived information including slope, aspect, and SRI (Shaded Relief Image). The experiments were carried out using DSM and label dataset provided by ISPRS (International Society for Photogrammetry and Remote Sensing) for CNN-based SegNet model. In particular, experiments focus on combining multi-source information to improve training performance and synergistic effect of the DL model. The results demonstrate that buildings were effectively classified and extracted by the proposed approach.

• The Journal of The Korea Institute of Intelligent Transport Systems
• /
• v.19 no.5
• /
• pp.70-81
• /
• 2020

Point-by-point data, which is abundantly collected by vehicles with embedded GPS (Global Positioning System), generate useful information. These data facilitate decisions by transportation jurisdictions, and private vendors can monitor and investigate micro-scale driver behavior, traffic flow, and roadway movements. The information is applied to develop app-based route guidance and business models. Of these, speed data play a vital role in developing key parameters and applying agent-based information and services. Nevertheless, link speed values require different levels of physical storage and fidelity, depending on both collecting and reporting intervals. Given these circumstances, this study aimed to establish an appropriate collection interval to efficiently utilize Space Mean Speed information by vehicles with embedded GPS. We conducted a comparison of Probe-vehicle data and Image-based vehicle data to understand PE(Percentage Error). According to the study results, the PE of the Probe-vehicle data showed a 95% confidence level within an 8-second interval, which was chosen as the appropriate collection interval for Probe-vehicle data. It is our hope that the developed guidelines facilitate C-ITS, and autonomous driving service providers will use more reliable Space Mean Speed data to develop better related C-ITS and autonomous driving services.

• Journal of Intelligence and Information Systems
• /
• v.29 no.3
• /
• pp.229-247
• /
• 2023

With the increasing emphasis on anomaly detection across various fields, diverse anomaly detection algorithms have been developed for various data types and anomaly patterns. However, the performance of anomaly detection algorithms is generally evaluated on publicly available datasets, and the specific performance of each algorithm on anomalies of particular types remains unexplored. Consequently, selecting an appropriate anomaly detection algorithm for specific analytical contexts poses challenges. Therefore, in this paper, we aim to investigate the types of anomalies and various attributes of data. Subsequently, we intend to propose approaches that can assist in the selection of appropriate anomaly detection algorithms based on this understanding. Specifically, this study compares the performance of anomaly detection algorithms for four types of anomalies: local, global, contextual, and clustered anomalies. Through further analysis, the impact of label availability, data quantity, and dimensionality on algorithm performance is examined. Experimental results demonstrate that the most effective algorithm varies depending on the type of anomaly, and certain algorithms exhibit stable performance even in the absence of anomaly-specific information. Furthermore, in some types of anomalies, the performance of unsupervised anomaly detection algorithms was observed to be lower than that of supervised and semi-supervised learning algorithms. Lastly, we found that the performance of most algorithms is more strongly influenced by the type of anomalies when the data quantity is relatively scarce or abundant. Additionally, in cases of higher dimensionality, it was noted that excellent performance was exhibited in detecting local and global anomalies, while lower performance was observed for clustered anomaly types.

• Journal of Internet Computing and Services
• /
• v.6 no.1
• /
• pp.85-93
• /
• 2005

This paper describes how to distribute high multi-dimensional facial expression data of vast quantity over a suitable space and produce facial expression animations by selecting expressions while animator navigates this space in real-time. We have constructed facial spaces by using about 2400 facial expression frames on this paper. These facial spaces are created by calculating of the shortest distance between two random expressions. The distance between two points In the space of expression, which is manifold space, is described approximately as following; When the linear distance of them is shorter than a decided value, if the two expressions are adjacent after defining the expression state vector of facial status using distance matrix expressing distance between two markers, this will be considered as the shortest distance (manifold distance) of the two expressions. Once the distance of those adjacent expressions was decided, We have taken a Floyd algorithm connecting these adjacent distances to yield the shortest distance of the two expressions. We have used CCA(Curvilinear Component Analysis) technique to visualize multi-dimensional spaces, the form of expressing space, into two dimensions. While the animators navigate this two dimensional spaces, they produce a facial animation by using user interface in real-time.