• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제13권4호
• /
• pp.2129-2147
• /
• 2019

This paper introduces an adaptive face recognition method based on a Novel Incremental Kernel Nonparametric Discriminant Analysis (IKNDA) that is able to learn through time. More precisely, the IKNDA has the advantage of incrementally reducing data dimension, in a discriminative manner, as new samples are added asynchronously. Thus, it handles dynamic and large data in a better way. In order to perform face recognition effectively, we combine the Gabor features and the ordinal measures to extract the facial features that are coded across local parts, as visual primitives. The variegated ordinal measures are extraught from Gabor filtering responses. Then, the histogram of these primitives, across a variety of facial zones, is intermingled to procure a feature vector. This latter's dimension is slimmed down using PCA. Finally, the latter is treated as a facial vector input for the advanced IKNDA. A comparative evaluation of the IKNDA is performed for face recognition, besides, for other classification endeavors, in a decontextualized evaluation schemes. In such a scheme, we compare the IKNDA model to some relevant state-of-the-art incremental and batch discriminant models. Experimental results show that the IKNDA outperforms these discriminant models and is better tool to improve face recognition performance.

• 한국컴퓨터정보학회논문지
• /
• 제24권3호
• /
• pp.143-150
• /
• 2019

This study aimed to analyze the keyword related to Health Disparity in Korea through the method of keyword network analysis and to establish a basic database for suggesting ideas for prospective studies in physical education. To achieve the goal, this study crawled co-occured keyword with 'health' and 'disparity' from news casted in 20 different channels. The duration of the news was 3 months, from September 11th, 2018 to December 11th. The results are as follows. First, among the news during recent 3 months, there were 1,383 keyword related to health disparity and this study selected 173 keyword which had co-occured over 3 times. Second, the inclusiveness of the network was 97.674% and the density was .038. Third, analyzing news related to health disparity, 'mortality' was the most co-occured keyword and 'disparity', 'reinforcement', 'the most', 'health', '6 times', 'Seoul', 'half', 'medicine', and 'local' were shown similarly. And common keyword in 4 centrality were 13 keyword. Lastly, by analyzing eigenvector centrality, significantly different result has shown. 'Disparity' was the most co-occured keyword. Based on this result, this study showed the necessity for reinforcing the public physical education in public education system in Korea. In order to achieve it, the field of physical education must look beyond present elite-focused physical education to public physical activity.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제15권4호
• /
• pp.1275-1292
• /
• 2021

As a fine-grained classification problem, aspect-level sentiment classification predicts the sentiment polarity for different aspects in context. To address this issue, researchers have widely used attention mechanisms to abstract the relationship between context and aspects. Still, it is difficult to effectively obtain a more profound semantic representation, and the strong correlation between local context features and the aspect-based sentiment is rarely considered. In this paper, a hybrid attention capsule network for aspect-level sentiment classification (ABASCap) was proposed. In this model, the multi-head self-attention was improved, and a context mask mechanism based on adjustable context window was proposed, so as to effectively obtain the internal association between aspects and context. Moreover, the dynamic routing algorithm and activation function in capsule network were optimized to meet the task requirements. Finally, sufficient experiments were conducted on three benchmark datasets in different domains. Compared with other baseline models, ABASCap achieved better classification results, and outperformed the state-of-the-art methods in this task after incorporating pre-training BERT.

• Journal of Information Processing Systems
• /
• 제17권6호
• /
• pp.1115-1126
• /
• 2021

Plant disease is one of the most irritating problems for agriculture growers. Thus, timely detection of plant diseases is of high importance to practical value, and corresponding measures can be taken at the early stage of plant diseases. Therefore, numerous researchers have made unremitting efforts in plant disease identification. However, this problem was not solved effectively until the development of artificial intelligence and big data technologies, especially the wide application of deep learning models in different fields. Since the symptoms of plant diseases mainly appear visually on leaves, computer vision and machine learning technologies are effective and rapid methods for identifying various kinds of plant diseases. As one of the fruits with the highest nutritional value, apple production directly affects the quality of life, and it is important to prevent disease intrusion in advance for yield and taste. In this study, an improved deep residual network is proposed for apple leaf disease identification in a novel way, a global residual connection is added to the original residual network, and the local residual connection architecture is optimized. Including that 1,977 apple leaf disease images with three categories that are collected in this study, experimental results show that the proposed method has achieved 98.74% top-1 accuracy on the test set, outperforming the existing state-of-the-art models in apple leaf disease identification tasks, and proving the effectiveness of the proposed method.

• 로봇학회논문지
• /
• 제16권2호
• /
• pp.112-121
• /
• 2021

Unlike a typical small-sized robot navigating in a free space, an autonomous vehicle has to travel in a designated road which has lanes to follow and traffic rules to obey. High-Definition (HD) maps, which include road markings, traffic signs, and traffic lights with high location accuracy, can help an autonomous vehicle avoid the need to detect such challenging road surroundings. With space constraints and a pre-built HD map, a new type of path planning algorithm can be conceived as a substitute for conventional grid-based path planning algorithms, which require substantial planning time to cover large-scale free space. In this paper, we propose an obstacle-avoiding, cost-based planning algorithm in a continuous space that aims to pursue a globally-planned path with the help of HD map information. Experimentally, the proposed algorithm is shown to outperform other state-of-the-art path planning algorithms in terms of computation complexity in a typical urban road setting, thereby achieving real-time performance and safe avoidance of obstacles.

• Nuclear Engineering and Technology
• /
• 제54권4호
• /
• pp.1495-1507
• /
• 2022

The improvement of thermal-hydraulic analysis techniques is essential to ensure the safety and reliability of nuclear power plants. The one-dimensional two-fluid model has been adopted in state-of-the-art thermal-hydraulic system codes. Current constitutive equations used in the system codes reach a mature level. Some exceptions are the partition method of wall friction in the momentum equation of the two-fluid model and the interfacial drag force model for a horizontal two-phase flow. This study is focused on deriving the partition method of wall friction in the momentum equation of the two-fluid model and modeling the interfacial drag force model for a horizontal bubbly flow. The one-dimensional momentum equation in the two-fluid model is derived from the local momentum equation. The derived one-dimensional momentum equation demonstrates that total wall friction should be apportioned to gas and liquid phases based on the phasic volume fraction, which is the same as that used in the SPACE code. The constitutive equations for the interfacial drag force are also identified. Based on the assessments, the Rassame-Hibiki correlation, Hibiki-Ishii correlation, Ishii-Zuber correlation, and Rassame-Hibiki correlation are recommended for computing the distribution parameter, interfacial area concentration, drag coefficient, and relative velocity covariance of a horizontal bubbly flow, respectively.

• Acta Via Serica
• /
• 제7권2호
• /
• pp.1-38
• /
• 2022

Many knotted carpets from western China and Central Asia from 700 BCE to 700 CE demonstrate strong resemblances to the Assyrian-Persian stone reliefs and Hellenistic-Romanmosaics. Egypt reveals the earliest samples in plain linen fiber and carpets with colorful designs in the mural paintings. The Egyptian wall decoration in faience shows the motifs of waterlily in two variations in the design, both of which appear in the Assyrian floor thresholds designs. The same waterlily or four-petal design in the Egyptian and Assyrian stonework also appear in the Pazyryk knotted carpet in the same manner. The Assyrian thresholds show a composition with a central area and borders in a square or rectangular overall design, which again parallels the Pazyryk carpet. The ideas of composition and motifs commonly seen in the Hellenistic and Roman mosaics appear in the carpets of Central Asia and western China. The similarities between the stone mosaic works and carpets demonstrate clearly that the remote places and peoples in ancient times learned from, and exchanged with, each other the ideas and artistic styles, and integrated to create some local artistic traditions, thanks to the roles the easy-to-carry textiles played throughout the Silk Road. The carpets played their roles: as practical material for daily use, and as carrier of cultural information through their designs.

• ETRI Journal
• /
• 제44권6호
• /
• pp.1004-1019
• /
• 2022

For autonomous driving without high-definition maps, we present a model capable of generating multiple plausible paths from egocentric images for autonomous vehicles. Our generative model comprises two neural networks: feature extraction network (FEN) and path generation network (PGN). The FEN extracts meaningful features from an egocentric image, whereas the PGN generates multiple paths from the features, given a driving intention and speed. To ensure that the paths generated are plausible and consistent with the intention, we introduce an attentive discriminator and train it with the PGN under a generative adversarial network framework. Furthermore, we devise an interaction model between the positions in the paths and the intentions hidden in the positions and design a novel PGN architecture that reflects the interaction model for improving the accuracy and diversity of the generated paths. Finally, we introduce ETRIDriving, a dataset for autonomous driving, in which the recorded sensor data are labeled with discrete high-level driving actions, and demonstrate the state-of-the-art performance of the proposed model on ETRIDriving in terms of accuracy and diversity.

• ETRI Journal
• /
• 제44권6호
• /
• pp.1034-1046
• /
• 2022

The learning-based multiview stereo (MVS) methods for three-dimensional (3D) reconstruction generally use 3D volumes for depth inference. The quality of the reconstructed depth maps and the corresponding point clouds is directly influenced by the spatial resolution of the 3D volume. Consequently, these methods produce point clouds with sparse local regions because of the lack of the memory required to encode a high volume of information. Here, we apply the atrous spatial pyramid pooling (ASPP) module in MVS methods to obtain dense feature maps with multiscale, long-range, contextual information using high receptive fields. For a given 3D volume with the same spatial resolution as that in the MVS methods, the dense feature maps from the ASPP module encoded with superior information can produce dense point clouds without a high memory footprint. Furthermore, we propose a 3D loss for training the MVS networks, which improves the predicted depth values by 24.44%. The ASPP module provides state-of-the-art qualitative results by constructing relatively dense point clouds, which improves the DTU MVS dataset benchmarks by 2.25% compared with those achieved in the previous MVS methods.

• 한국비블리아학회지
• /
• 제18권2호
• /
• pp.183-202
• /
• 2007

본 연구의 목적은 농산어촌 지역에 학교도서관을 설치하여 효과적으로 운영하기 위해 농산어촌 지역의 학교도서관을 담당하는 교사들이 가지고 있는 학교도서관 운영에 대한 인식을 조사하는데 있다. 사용된 연구방법은 설문지법과 면담법이 사용되었다. 조사결과 농산어촌 학교도서관의 역할은 독서교육 공간과 학생들의 문화 및 여가활용 공간으로 나타났다. 특히 학생들이 사회생활에 필요한 견문을 넓히는 공간으로 다양한 정보매체 활용을 위한 공간을 강조하였다. 또한 농산어촌의 학교도서관은 지역주민과 결합된 문화공간의 기능을 강조하였다. 이에 본 연구에서는 유연성 있는 도서관 운영시간, 학업능력 향상 프로그램 교양능력향상 프로그램, 문화공간 프로그램을 제시하였다.