• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.11 no.10
• /
• pp.4968-4986
• /
• 2017

In this paper, we primarily address the difficulty of automatic generation of a plausible depth map from a single image in an unstructured environment. The aim is to extrapolate a depth map with a more correct, rich, and distinct depth order, which is both quantitatively accurate as well as visually pleasing. Our technique, which is fundamentally based on a preexisting DepthTransfer algorithm, transfers depth information at the level of superpixels. This occurs within a framework that replaces a pixel basis with one of instance-based learning. A vital superpixels feature enhancing matching precision is posterior incorporation of predictive semantic labels into the depth extraction procedure. Finally, a modified Cross Bilateral Filter is leveraged to augment the final depth field. For training and evaluation, experiments were conducted using the Make3D Range Image Dataset and vividly demonstrate that this depth estimation method outperforms state-of-the-art methods for the correlation coefficient metric, mean log10 error and root mean squared error, and achieves comparable performance for the average relative error metric in both efficacy and computational efficiency. This approach can be utilized to automatically convert 2D images into stereo for 3D visualization, producing anaglyph images that are visually superior in realism and simultaneously more immersive.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.11 no.4
• /
• pp.2058-2074
• /
• 2017

Dense Heterogeneous Cellular Networks(HCNs) offer a promising approach to meet the target of 1000x increase in aggregate data rates in 5G wireless communication systems. However how to best utilize the available radio resources at densely deployed small cells remains an open problem as those small cells are typically unplanned. In this paper we focus on balancing loads across macro cells and small cells by offloading users to small cells, as well as dynamically switching off underutilized small cells. We propose a joint user association and base station(BS) sleep mangement(UA-BSM) scheme that proactively offloads users to a fraction of the densely deployed small cells. We propose a heuristic algorithm that iteratively solves the user association problem and puts BSs with low loads into sleep. An interference relation matrix(IRM) is constructed to help us identify the candidate BSs that can be put into sleep. User associations are then aggregated to selected small cells that remain active. Simulation results show that our proposed approach achieves load balancing across macro and small cells and reduces the number of active BSs. Numerical results show user signal to interference ratio(SINR) can be improved by small cell sleep control.

• Journal of Power Electronics
• /
• v.18 no.1
• /
• pp.225-233
• /
• 2018

This paper proposes a universal system control strategy for voltage source converter (VSC) based high voltage direct current (HVDC) systems. The framework of the designed control strategy consists of five layer structures considering the topology and control characteristics of the VSC-HVDC system. The control commands sent from the topmost layer can be transmitted to the next layer based on the existing communication system. When the commands are sent to each substation, the following transmission of commands between the four lower layers are realized using the internal communication system while ignoring the communication delay. This hierarchical control strategy can be easily applied to any VSC-HVDC system with any topology. Furthermore, an integrated controller for each converter is designed and implemented considering all of the possible operating states. The modular-designed integrated controller makes it quite easy to extend its operating states if necessary, and it is available for any kind of VSC. A detailed model of a VSC-HVDC system containing a DC hub is built in the PSCAD/EMTDC environment. Simulation results based on three operating conditions (the start-up process, the voltage margin control method and the master-slave control method) demonstrate the flexibility and validity of the proposed control strategy.

• Journal of Ginseng Research
• /
• v.45 no.4
• /
• pp.473-481
• /
• 2021

Background: Mitochondrial dysfunction is one of the significant reasons for Alzheimer's disease (AD). Ginsenosides, natural molecules extracted from Panax ginseng, have been demonstrated to exert essential neuroprotective functions, which can ascribe to its anti-oxidative effect, enhancing central metabolism and improving mitochondrial function. However, a comprehensive analysis of cellular mitochondrial bioenergetics after ginsenoside treatment under Aβ-oxidative stress is missing. Methods: The antioxidant activities of ginsenoside Rb₁, Rd, Re, Rg₁ were compared by measuring the cell survival and reactive oxygen species (ROS) formation. Next, the protective effects of ginsenosides of mitochondrial bioenergetics were examined by measuring oxygen consumption rate (OCR) in PC12 cells under Aβ-oxidative stress with an extracellular flux analyzer. Meanwhile, mitochondrial membrane potential (MMP) and mitochondrial dynamics were evaluated by confocal laser scanning microscopy. Results: Ginsenoside Rg₁ possessed the strongest anti-oxidative property, and which therefore provided the best protective function to PC12 cells under the Aβ oxidative stress by increasing ATP production to 3 folds, spare capacity to 2 folds, maximal respiration to 2 folds and non-mitochondrial respiration to 1.5 folds, as compared to Aβ cell model. Furthermore, ginsenoside Rg1 enhanced MMP and mitochondrial interconnectivity, and simultaneously reduced mitochondrial circularity. Conclusion: In the present study, these results demonstrated that ginsenoside Rg₁ could be the best natural compound, as compared with other ginsenosides, by modulating the OCR of cultured PC12 cells during oxidative phosphorylation, in regulating MMP and in improving mitochondria dynamics under Aβ-induced oxidative stress.

• Asian Pacific Journal of Cancer Prevention
• /
• v.15 no.24
• /
• pp.11001-11001
• /
• 2015

• Journal of Communications and Networks
• /
• v.18 no.4
• /
• pp.649-657
• /
• 2016

In this paper, we investigate energy efficiency (EE) of the traditional co-located and the distributed massive multiple-input multiple-output (MIMO) systems. First, we derive an approximate EE expression for both the idealistic and the realistic power consumption models. Then an optimal energy-efficient remote access unit (RAU) selection algorithm based on the distance between the mobile stations (MSs) and the RAUs are developed to maximize the EE for the downlink distributed massive MIMO systems under the realistic power consumption model. Numerical results show that the EE of the distributed massive MIMO systems is larger than the co-located massive MIMO systems under both the idealistic and realistic power consumption models, and the optimal EE can be obtained by the developed energy-efficient RAU selection algorithm.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.14 no.11
• /
• pp.4557-4572
• /
• 2020

The Alaska competition provides an opportunity to study the practical problems of real-world steganalysis. Participants are required to solve steganalysis involving various embedding schemes, inconsistency JPEG Quality Factor and various processing pipelines. In this paper, we propose a method to ensemble multiple deep learning steganalyzers. We select SRNet and RESDET as our base models. Then we design a three-layers model ensemble network to fuse these base models and output the final prediction. By separating the three colors channels for base model training and feature replacement strategy instead of simply merging features, the performance of the model ensemble is greatly improved. The proposed method won second place in the Alaska 1 competition in the end.

• International conference on construction engineering and project management
• /
• 2017.10a
• /
• pp.353-361
• /
• 2017

In order to improve the urban ecological environment, the central government has developed a series of water pollution control policies and measures since April 2015, and required local governments to complete the work of the polluted river and lake restoration within specified period. Moreover, the polluted river and lake restoration has been selected as a key evaluation indicator of achievements of the sponge city construction implemented since April 2015. This paper describes how to apply the sponge city construction technology to rehabilitate the polluted rivers and lakes through the polluted water treatment project in Xinghan New Area, Hanzhong and the polluted Huaguping River restoration project in Pingshan District, Shenzhen. The experience and lessons of the polluted river and lake restoration are summed up. Some measures to solve existing problems are put forward.

• International conference on construction engineering and project management
• /
• 2017.10a
• /
• pp.346-352
• /
• 2017

This study discusses application of the storm water management model (SWMM) to designing the sponge city facilities in the Athletes Village of Military World Games in Wuhan in October 2019. The SWMM was used to simulate the runoff processes and reduction efficiencies of the sponge city facilities. The runoffs of the sponge city facilities were compared with those of traditional drainage system for the design rainfall of 35.2mm and the rainfalls with different recurrence periods. The results show that the hign density sponge city facilities could meet the requirements for 80% of annual runoff control rate, SWMM can determine the scales of the sponge city facilities and effectively simulate the hydrological processes for different layout schemes. The simulation model is also helpful to making optimization of the sponge city facility layout.

• Journal of the Korean Society of Clothing and Textiles
• /
• v.29 no.1 s.139
• /
• pp.124-135
• /
• 2005

This study was intended to compare and analyze the clothing purchasing behavior of Chinese career women in Benjing, Shanghai, Shenzhen, and Changchun, and to provide useful information for establishment of marketing strategies for China. The subjects were 662 career women in their 20's or 30's living in Benjing, Shanghai, Shenzhen, and Changchun. Data was collected through the questionnaire based on the theoretical and previous studies from April 9, 2003 to April 31, 2003. A total of 662 questionnaires was analyzed by frequency, percentage, ANOVA, Duncan's multiple range test, and ${\chi}^2-test$. The results of this study were as follows: 1. Demographic variables such as an age, marriage, academic back ground, residential city, monthly income, and monthly expenditure on clothing were significantly different among the career women in Benjing, Shanghai, Shenzhen, and Changchun. 2. Clothing purchasing behavior variables such as clothing purchasing motive, using informants, clothing selection standards, store selection standards, and satisfaction after purchasing clothes were significantly different among the career women in Benjing, Shanghai, Shenzhen, and Changchun.