• Journal of the Korean Society for Precision Engineering
• /
• v.21 no.12
• /
• pp.100-108
• /
• 2004

In reverse engineering, data acquisition methodology can generally be categorized into contacting and non-contacting types. Recently, researches on hybrid or sensor fusion of the two types have been increasing. In addition, efficient construction of a geometric model from the measurement data is required, where considerable amount of user interaction to classify and localize regions of interest is inevitable. Our research focuses on the classification of each bounded region into a pre-defined feature shape fer a hybrid measuring scheme, where the overall procedures are described as fellows. Firstly, the physical model is digitized by a non-contacting laser scanner which rapidly provides cloud-of-points data. Secondly, the overall digitized data are approximated to a z-map model. Each bounding curve of a region of interest (featured area) can be 1.aced out based on our previous research. Then each confined area is systematically classified into one of the pre-defined feature types such as floor, wall, strip or volume, followed by a more accurate measuring step via a contacting probe. Assigned to each feature is a specific digitizing path topology which may reflect its own geometric character. The research can play an important role in minimizing user interaction at the stage of digitizing path planning.

• Journal of the Korean Operations Research and Management Science Society
• /
• v.39 no.4
• /
• pp.75-84
• /
• 2014

In this paper, we dealt with feature selection problem of large-scale and high-dimensional biological data such as omics data. For this problem, most of the previous approaches used simple score function to reduce the number of original variables and selected features from the small number of remained variables. In the case of methods that do not rely on filtering techniques, they do not consider the interactions between the variables, or generate approximate solutions to the simplified problem. Unlike them, by combining set covering and clustering techniques, we developed a new method that could deal with total number of variables and consider the combinatorial effects of variables for selecting good features. To demonstrate the efficacy and effectiveness of the method, we downloaded gene expression datasets from TCGA (The Cancer Genome Atlas) and compared our method with other algorithms including WEKA embeded feature selection algorithms. In the experimental results, we showed that our method could select high quality features for constructing more accurate classifiers than other feature selection algorithms.

• Proceedings of the KSRS Conference
• /
• 2004.10a
• /
• pp.359-362
• /
• 2004

In multi-dimensional image, ICA-based feature extraction algorithm, which is proposed in this paper, is for the purpose of detecting target feature about pixel assumed as a linear mixed spectrum sphere, which is consisted of each different type of material object (target feature and background feature) in spectrum sphere of reflectance of each pixel. Landsat ETM+ satellite image is consisted of multi-dimensional data structure and, there is target feature, which is purposed to extract and various background image is mixed. In this paper, in order to eliminate background features (tidal flat, seawater and etc) around target feature (laver farm) effectively, pixel spectrum sphere of target feature is projected onto the orthogonal spectrum sphere of background feature. The rest amount of spectrum sphere of target feature in the pixel can be presumed to remove spectrum sphere of background feature. In order to make sure the excellence of feature extraction method based on ICA, which is proposed in this paper, laver farm feature extraction from Landsat ETM+ satellite image is applied. Also, In the side of feature extraction accuracy and the noise level, which is still remaining not to remove after feature extraction, we have conducted a comparing test with traditionally most popular method, maximum-likelihood. As a consequence, the proposed method from this paper can effectively eliminate background features around mixed spectrum sphere to extract target feature. So, we found that it had excellent detection efficiency.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.11 no.9
• /
• pp.4527-4548
• /
• 2017

This paper proposes a method to generate panoramic images by combining conventional feature extraction algorithms (e.g., SIFT, SURF, MPEG-7 CDVS) with sensed data from inertia sensors to enhance the stitching results. The challenge of image stitching increases when the images are taken from two different mobile phones with no posture calibration. Using inertia sensor data obtained by the mobile phone, images with different yaw, pitch, and roll angles are preprocessed and adjusted before performing stitching process. Performance of stitching (e.g., feature extraction time, inlier point numbers, stitching accuracy) between conventional feature extraction algorithms is reported along with the stitching performance with/without using the inertia sensor data. In addition, the stitching accuracy of video data was improved using the same sensed data, with discrete calculation of homograph matrix. The experimental results for stitching accuracies and speed using sensed data are presented in this paper.

• KIPS Transactions on Software and Data Engineering
• /
• v.10 no.9
• /
• pp.375-384
• /
• 2021

Deep learning techniques have been proven to have high performance in image processing and are applied in various fields. The most widely used methods for validating a deep learning model include a holdout verification method, a k-fold cross verification method, and a bootstrap method. These legacy methods consider the balance of the ratio between classes in the process of dividing the data set, but do not consider the ratio of various features that exist within the same class. If these features are not considered, verification results may be biased toward some features. Therefore, we propose a deep learning model validation method based on data feature coverage for image classification by improving the legacy methods. The proposed technique proposes a data feature coverage that can be measured numerically how much the training data set for training and validation of the deep learning model and the evaluation data set reflects the features of the entire data set. In this method, the data set can be divided by ensuring coverage to include all features of the entire data set, and the evaluation result of the model can be analyzed in units of feature clusters. As a result, by providing feature cluster information for the evaluation result of the trained model, feature information of data that affects the trained model can be provided.

• Geomechanics and Engineering
• /
• v.25 no.1
• /
• pp.1-16
• /
• 2021

Machine learning models have been widely used for landslide susceptibility assessment (LSA) in recent years. The large number of inputs or conditioning factors for these models, however, can reduce the computation efficiency and increase the difficulty in collecting data. Feature selection is a good tool to address this problem by selecting the most important features among all factors to reduce the size of the input variables. However, two important questions need to be solved: (1) how do feature selection methods affect the performance of machine learning models? and (2) which feature selection method is the most suitable for a given machine learning model? This paper aims to address these two questions by comparing the predictive performance of 13 feature selection-based machine learning (FS-ML) models and 5 ordinary machine learning models on LSA. First, five commonly used machine learning models (i.e., logistic regression, support vector machine, artificial neural network, Gaussian process and random forest) and six typical feature selection methods in the literature are adopted to constitute the proposed models. Then, fifteen conditioning factors are chosen as input variables and 1,017 landslides are used as recorded data. Next, feature selection methods are used to obtain the importance of the conditioning factors to create feature subsets, based on which 13 FS-ML models are constructed. For each of the machine learning models, a best optimized FS-ML model is selected according to the area under curve value. Finally, five optimal FS-ML models are obtained and applied to the LSA of the studied area. The predictive abilities of the FS-ML models on LSA are verified and compared through the receive operating characteristic curve and statistical indicators such as sensitivity, specificity and accuracy. The results showed that different feature selection methods have different effects on the performance of LSA machine learning models. FS-ML models generally outperform the ordinary machine learning models. The best FS-ML model is the recursive feature elimination (RFE) optimized RF, and RFE is an optimal method for feature selection.

• Journal of Broadcast Engineering
• /
• v.27 no.3
• /
• pp.283-294
• /
• 2022

The feature map used in the network for deep learning generally has larger data than the image and a higher compression rate than the image compression rate is required to transmit the feature map. This paper proposes a method for transmitting a pyramid feature map with high compression rate, which is used in a network with an FPN structure that has robustness to object size in deep learning-based image processing. In order to efficiently compress the pyramid feature map, this paper proposes a structure that predicts a pyramid feature map of a level that is not transmitted with pyramid feature map of some levels that transmitted through the proposed prediction network to efficiently compress the pyramid feature map and restores compression damage through the proposed reconstruction network. Suggested mAP, the performance of object detection for the COCO data set 2017 Train images of the proposed method, showed a performance improvement of 31.25% in BD-rate compared to the result of compressing the feature map through VTM12.0 in the rate-precision graph, and compared to the method of performing compression through PCA and DeepCABAC, the BD-rate improved by 57.79%.

• The Journal of the Acoustical Society of Korea
• /
• v.24 no.4
• /
• pp.210-216
• /
• 2005

We Propose a new feature normalization scheme based on eigenspace for achieving robust speech recognition. In general, mean and variance normalization (MVN) is Performed in cepstral domain. However, another MVN approach using eigenspace was recently introduced. in that the eigenspace normalization Procedure Performs normalization in a single eigenspace. This Procedure consists of linear PCA matrix feature transformation followed by mean and variance normalization of the transformed cepstral feature. In this method. 39 dimensional feature distribution is represented using only a single eigenspace. However it is observed to be insufficient to represent all data distribution using only a sin91e eigenvector. For more specific representation. we apply unique na independent eigenspaces to cepstra, delta and delta-delta cepstra respectively in this Paper. We also normalize training data in eigenspace and get the model from the normalized training data. Finally. a feature space rotation procedure is introduced to reduce the mismatch of training and test data distribution in noisy condition. As a result, we obtained a substantial recognition improvement over the basic eigenspace normalization.

• Journal of the Institute of Electronics Engineers of Korea CI
• /
• v.47 no.3
• /
• pp.36-41
• /
• 2010

An advanced form of Multistage Feature-based Classification Model(AMFCM), called AMFCM, is proposed in this paper. AMFCM like MFCM does not use the concatenated form of available feature vectors extracted from original data to classify each data, but uses only groups related to each feature vector to classify separately. The prpposed AMFCM improves the contribution rate used in MFCM and proposes a confusion table for each local classifier using a specific feature vector group. The confusion table for each local classifier contains accuracy information of each local classifier on each class of data. The proposed AMFCM is applied to the problem of music genre classification on a set of music data. The results demonstrate that the proposed AMFCM outperforms MFCM by 8% - 15% on average in terms of classification accuracy depending on the grouping algorithms used for local classifiers and the number of clusters.

• Journal of the Korea Society of Computer and Information
• /
• v.29 no.3
• /
• pp.55-66
• /
• 2024

Gender classification techniques have received a lot of attention from researchers because they can be used in various fields such as forensics, surveillance systems, and demographic studies. As previous studies have shown that there are distinctive features between male and female gait, various techniques have been proposed to classify gender from three dimensional(3-D) gait data. However, some of the gait features extracted from 3-D gait data using existing techniques are similar or redundant to each other or do not help in gender classification. In this study, we propose a method to select features that are useful for gender classification using a correlation-based feature selection technique. To demonstrate the effectiveness of the proposed feature selection technique, we compare the performance of gender classification models before and after applying the proposed feature selection technique using a 3-D gait dataset available on the Internet. Eight machine learning algorithms applicable to binary classification problems were utilized in the experiments. The experimental results show that the proposed feature selection technique can reduce the number of features by 22, from 82 to 60, while maintaining the gender classification performance.