• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.17 no.8
• /
• pp.1996-2015
• /
• 2023

High-definition (HD) maps can provide precise road information that enables an autonomous driving system to effectively navigate a vehicle. Recent research has focused on leveraging semantic segmentation to achieve automatic annotation of HD maps. However, the existing methods suffer from low recognition accuracy in automatic driving scenarios, leading to inefficient annotation processes. In this paper, we propose a novel semantic segmentation method for automatic HD map annotation. Our approach introduces a new encoder, known as the convolutional transformer hybrid encoder, to enhance the model's feature extraction capabilities. Additionally, we propose a multi-level fusion module that enables the model to aggregate different levels of detail and semantic information. Furthermore, we present a novel decoupled boundary joint decoder to improve the model's ability to handle the boundary between categories. To evaluate our method, we conducted experiments using the Bird's Eye View point cloud images dataset and Cityscapes dataset. Comparative analysis against stateof-the-art methods demonstrates that our model achieves the highest performance. Specifically, our model achieves an mIoU of 56.26%, surpassing the results of SegFormer with an mIoU of 1.47%. This innovative promises to significantly enhance the efficiency of HD map automatic annotation.

• Communications for Statistical Applications and Methods
• /
• v.30 no.6
• /
• pp.589-603
• /
• 2023

Deep generative models target to infer the underlying true data distribution, and it leads to a huge success in generating fake-but-realistic data. Regarding such a perspective, the data attributes can be a crucial factor in the data generation process since non-existent counterfactual samples can be generated by altering certain factors. For example, we can generate new portrait images by flipping the gender attribute or altering the hair color attributes. This paper proposes counterfactual disentangled variational autoencoder generative adversarial networks (CDVAE-GAN), specialized for data attribute level counterfactual data generation. The structure of the proposed CDVAE-GAN consists of variational autoencoders and generative adversarial networks. Specifically, we adopt a Gaussian variational autoencoder to extract low-dimensional disentangled data features and auxiliary Bernoulli latent variables to model the data attributes separately. Also, we utilize a generative adversarial network to generate data with high fidelity. By enjoying the benefits of the variational autoencoder with the additional Bernoulli latent variables and the generative adversarial network, the proposed CDVAE-GAN can control the data attributes, and it enables producing counterfactual data. Our experimental result on the CelebA dataset qualitatively shows that the generated samples from CDVAE-GAN are realistic. Also, the quantitative results support that the proposed model can produce data that can deceive other machine learning classifiers with the altered data attributes.

• Journal of the Korean Institute of Illuminating and Electrical Installation Engineers
• /
• v.16 no.6
• /
• pp.66-73
• /
• 2002

This paper presents ZVZCS(Zero-Voltage and Zero-Current Switching) Three Level DC/DC Converter without primary freewheeling diodes. The new converter presented in this paper used a phase shirt control with a flying capacitor in the primary side to achieve ZVS for the outer switches. A secondary anxiliary circuit which consists of one small capacitor, two small diodes and one coupled inductor, is added in the secondary to provide ZVZCS conditions to primary switches, ZVS for outer switches and ZCS for inner switches. Many advantages include simple secondary auxiliary circuit topology, high efficiency, and low cost make the new converter attractive for high power applications. Also the circulating current flows through the circuit so that it causes the needless coduction loss to be occurred in the devices and the transformer of the circuit The new converter has no primary auxiliary diodes for freewheeling current. The principle of operation, feature and design considerations are illustrated and verified through the experiment with a 1[㎾］ 50[KHz］IGBT based experimental circuit.

• Journal of the Institute of Electronics Engineers of Korea CI
• /
• v.37 no.6
• /
• pp.22-31
• /
• 2000

In this paper, we propose a method for removing noise components in the feature extraction process for robust speech recognition. This method is based on blind separation using independent component analysis (ICA). Given two noisy speech recordings the algorithm linearly separates speech from the unwanted noise signal. To apply ICA as closely as possible to the feature level for recognition, a new spectral analysis is presented. It modifies the computation of band energies by previously averaging out fast Fourier transform (FFT) points in several divided ranges within one met-scaled band. The simple analysis using sample variances of band energies of speech and noise, and recognition experiments showed its noise robustness. For noisy speech signals recorded in real environments, the proposed method which applies ICA to the new spectral analysis improved the recognition performances to a considerable extent, and was particularly effective for low signal-to-noise ratios (SNRs). This method gives some insights into applying ICA to feature levels and appears useful for robust speech recognition.

• Journal of Animal Science and Technology
• /
• v.63 no.2
• /
• pp.367-379
• /
• 2021

The objectives of this study were to evaluate convolutional neural network models and computer vision techniques for the classification of swine posture with high accuracy and to use the derived result in the investigation of the effect of dietary fiber level on the behavioral characteristics of the pregnant sow under low and high ambient temperatures during the last stage of gestation. A total of 27 crossbred sows (Yorkshire × Landrace; average body weight, 192.2 ± 4.8 kg) were assigned to three treatments in a randomized complete block design during the last stage of gestation (days 90 to 114). The sows in group 1 were fed a 3% fiber diet under neutral ambient temperature; the sows in group 2 were fed a diet with 3% fiber under high ambient temperature (HT); the sows in group 3 were fed a 6% fiber diet under HT. Eight popular deep learning-based feature extraction frameworks (DenseNet121, DenseNet201, InceptionResNetV2, InceptionV3, MobileNet, VGG16, VGG19, and Xception) used for automatic swine posture classification were selected and compared using the swine posture image dataset that was constructed under real swine farm conditions. The neural network models showed excellent performance on previously unseen data (ability to generalize). The DenseNet121 feature extractor achieved the best performance with 99.83% accuracy, and both DenseNet201 and MobileNet showed an accuracy of 99.77% for the classification of the image dataset. The behavior of sows classified by the DenseNet121 feature extractor showed that the HT in our study reduced (p < 0.05) the standing behavior of sows and also has a tendency to increase (p = 0.082) lying behavior. High dietary fiber treatment tended to increase (p = 0.064) lying and decrease (p < 0.05) the standing behavior of sows, but there was no change in sitting under HT conditions.

• Women's Studies Review
• /
• v.29 no.2
• /
• pp.139-173
• /
• 2012

This paper examines the delayed situations for gender equality in South Korean, Japanese, and Taiwanese families despite the challenge to the gender division of labor in modern society, and to analyze the contradiction between the notions of gender equality and the experiences women face in East Asia countries. Using EASS data, we analyze the effective difference over the division of household labor according to women's age and length of school time, attitude for gender division of labor, couple's labor time, and family network. In South Korea and Taiwan, men's actual ratio of household division is higher than Japanese men's. On the other hand, Japanese women's ratio of household division is the highest in spite of their progressive attitude for gender equality. It is due to the difference of women's working time among the countries. In South Korea and Taiwan, women tend to work in full time job, so that they seem to inevitably reduce the time for household labor. The family characteristics have an effect on the women's ratio of household division in Taiwan, and the feature of women's employment does in South Korea. The high percentage of three-generation household contributes to the reduction of housework burden in Taiwan. In South Korea, the higher women's education levels, the higher the women's ratio of household division. Women's weakened bargaining power for household labor is due to the relatively low level of high-educated women's economic participation in South Korea. This paper reveals the effective factors on the gender division of household labor. We propose the necessity of the macro-level analysis as well as the analysis of the personal and conjugal feature.

• Journal of the Korean Society of Radiology
• /
• v.3 no.1
• /
• pp.29-39
• /
• 2009

In this paper has proposed to the recognition of the disease on medical images using neural network. The neural network is constructed as three-layers of the input-layer, the hidden-layer and the output-layer. The training method applied for the recognition of disease region is adaptive error back-propagation. The low-frequency region analyzed by DWT are expressed by matrix. The coefficient-values of the characteristic polynomial applied are n+1. The normalized maximum value +1 and minimum value -1 in the range of tangent-sigmoid transfer function are applied to be use as the input vector of the neural network. To prove the validity of the proposed methods used in the experiment with a simulation experiment, the input medical image recognition rate the evaluation of areas of disease. As a result of the experiment, the characteristic polynomial coefficient of low-frequency area matrix, conversed to 4 level DWT, was proved to be optimum to be applied to the feature parameter. As for the number of training, it was marked fewest in 0.01 of learning coefficient and 0.95 of momentum, when the adaptive error back-propagation was learned by inputting standardized feature parameter into organized neural network. As to the training result when the learning coefficient was 0.01, and momentum was 0.95, it was 100% recognized in fifty-five times of the stomach image, fifty-five times of the chest image, forty-six times of the CT image, fifty-five times of ultrasonogram, and one hundred fifty-seven times of angiogram.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.20 no.10
• /
• pp.1-6
• /
• 2019

As mobile technologies such as Internet of Things (IoT)-based smart homes and financial technologies (FinTech) are developed, authentication by smart devices is used everywhere. As a result, presence-based biometric authentication using smart devices has become a new mainstream in knowledge-based authentication methods like the existing passwords. The electrocardiogram (ECG) is less prone to forgery, and high-level personal identification is its unique feature from among various biometric authentication methods, such as the pulse, fingerprints, the face, and the iris. Biometric authentication using an ECG is receiving a great deal of attention due to its uses in healthcare and FinTech. In this study, we implemented an ECG authentication system that allows users to easily measure and authenticate their ECG waveforms using a miniaturized wearable device, rather than a large and expensive measurement device. The implemented ECG authentication system identifies ECG features through P-Q-R-S-T feature point identification, and was user-certified under the proposed authentication protocols. Finally, assessment of measurements in a majority of adult males showed a relatively low false acceptance rate of 1.73%, and a low false rejection rate of 4.14%, in a stable normal state. In a high-activity state, the false acceptance rate was 13.72%, and the false rejection rate was 21.68%. In a high-heart rate state, the false acceptance rate was 10.48%, and the false rejection rate was 11.21%.

• Journal of Intelligence and Information Systems
• /
• v.26 no.4
• /
• pp.173-198
• /
• 2020

For a long time, many studies have been conducted on predicting the success of campaigns for customers in academia, and prediction models applying various techniques are still being studied. Recently, as campaign channels have been expanded in various ways due to the rapid revitalization of online, various types of campaigns are being carried out by companies at a level that cannot be compared to the past. However, customers tend to perceive it as spam as the fatigue of campaigns due to duplicate exposure increases. Also, from a corporate standpoint, there is a problem that the effectiveness of the campaign itself is decreasing, such as increasing the cost of investing in the campaign, which leads to the low actual campaign success rate. Accordingly, various studies are ongoing to improve the effectiveness of the campaign in practice. This campaign system has the ultimate purpose to increase the success rate of various campaigns by collecting and analyzing various data related to customers and using them for campaigns. In particular, recent attempts to make various predictions related to the response of campaigns using machine learning have been made. It is very important to select appropriate features due to the various features of campaign data. If all of the input data are used in the process of classifying a large amount of data, it takes a lot of learning time as the classification class expands, so the minimum input data set must be extracted and used from the entire data. In addition, when a trained model is generated by using too many features, prediction accuracy may be degraded due to overfitting or correlation between features. Therefore, in order to improve accuracy, a feature selection technique that removes features close to noise should be applied, and feature selection is a necessary process in order to analyze a high-dimensional data set. Among the greedy algorithms, SFS (Sequential Forward Selection), SBS (Sequential Backward Selection), SFFS (Sequential Floating Forward Selection), etc. are widely used as traditional feature selection techniques. It is also true that if there are many risks and many features, there is a limitation in that the performance for classification prediction is poor and it takes a lot of learning time. Therefore, in this study, we propose an improved feature selection algorithm to enhance the effectiveness of the existing campaign. The purpose of this study is to improve the existing SFFS sequential method in the process of searching for feature subsets that are the basis for improving machine learning model performance using statistical characteristics of the data to be processed in the campaign system. Through this, features that have a lot of influence on performance are first derived, features that have a negative effect are removed, and then the sequential method is applied to increase the efficiency for search performance and to apply an improved algorithm to enable generalized prediction. Through this, it was confirmed that the proposed model showed better search and prediction performance than the traditional greed algorithm. Compared with the original data set, greed algorithm, genetic algorithm (GA), and recursive feature elimination (RFE), the campaign success prediction was higher. In addition, when performing campaign success prediction, the improved feature selection algorithm was found to be helpful in analyzing and interpreting the prediction results by providing the importance of the derived features. This is important features such as age, customer rating, and sales, which were previously known statistically. Unlike the previous campaign planners, features such as the combined product name, average 3-month data consumption rate, and the last 3-month wireless data usage were unexpectedly selected as important features for the campaign response, which they rarely used to select campaign targets. It was confirmed that base attributes can also be very important features depending on the type of campaign. Through this, it is possible to analyze and understand the important characteristics of each campaign type.

• Journal of Korea Multimedia Society
• /
• v.7 no.9
• /
• pp.1312-1320
• /
• 2004

Recently, even if a lot of visual feature representations have been studied and systems have been built, there is a limit to existing content-based image retrieval mechanism in its availability. One of the limits is the gap between a user's high-level concepts and a system's low-level features. And human beings' subjectivity in perceiving similarity is excluded. Therefore, correct visual information delivery and a method that can retrieve the data efficiently are required. Relevance feedback can increase the efficiency of image retrieval because it responds of a user's information needs in multimedia retrieval. This paper proposes an efficient CBIR introducing positive and negative relevance feedback with threshold value and pre-fetching to improve the performance of conventional relevance feedback mechanisms. With this Proposed feedback strategy, we implement an image retrieval system that improves the conventional retrieval system.