• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.18 no.4
• /
• pp.938-958
• /
• 2024

Unsupervised Video Object Segmentation (UVOS) is a highly challenging problem in computer vision as the annotation of the target object in the testing video is unknown at all. The main difficulty is to effectively handle the complicated and changeable motion state of the target object and the confusion of similar background objects in video sequence. In this paper, we propose a novel deep Dual-stream Co-enhanced Network (DC-Net) for UVOS via bidirectional motion cues refinement and multi-level feature aggregation, which can fully take advantage of motion cues and effectively integrate different level features to produce high-quality segmentation mask. DC-Net is a dual-stream architecture where the two streams are co-enhanced by each other. One is a motion stream with a Motion-cues Refine Module (MRM), which learns from bidirectional optical flow images and produces fine-grained and complete distinctive motion saliency map, and the other is an appearance stream with a Multi-level Feature Aggregation Module (MFAM) and a Context Attention Module (CAM) which are designed to integrate the different level features effectively. Specifically, the motion saliency map obtained by the motion stream is fused with each stage of the decoder in the appearance stream to improve the segmentation, and in turn the segmentation loss in the appearance stream feeds back into the motion stream to enhance the motion refinement. Experimental results on three datasets (Davis2016, VideoSD, SegTrack-v2) demonstrate that DC-Net has achieved comparable results with some state-of-the-art methods.

• Journal of IKEEE
• /
• v.23 no.4
• /
• pp.1366-1372
• /
• 2019

In this paper, we designed a revised neural network to remove the Monte Carlo Rendering noise contained in the ray tracing graphics. The Monte Carlo Rendering is the best way to enhance the graphic's realism, but because of the need to calculate more than thousands of light effects per pixel, rendering processing time has increased rapidly, causing a major problem with real-time processing. To improve this problem, the number of light used in pixels is reduced, where rendering noise occurs and various studies have been conducted to eliminate this noise. In this paper, a deep learning is used to remove rendering noise, especially by separating the rendering image into diffuse and specular light, so that the structure of the dual neural network is designed. As a result, the dual neural network improved by an average of 0.58 db for 64 test images based on PSNR, and 99.22% less light compared to reference image, enabling real-time race-tracing rendering.

• Journal of the Institute of Electronics and Information Engineers
• /
• v.53 no.10
• /
• pp.40-47
• /
• 2016

In this paper, a dual mode buck converter with an ability to change mode is proposed, which is suitable particularly for portable device. The problem of conventional mode control circuit is affected by load variation condition such as suddenly or slowly. To resolve this problem, the mode control was designed with slow clock method. Also, when change from the PFM(Pulse Frequency Modulation) mode to the PWM(Pulse Width Modulation) mode, to use the counter to detect a high load. And the user can select mode transition point in load range from 20mA to 90mA by 3 bit digital signal. The circuits are implemented by using BCDMOS 0.18um 2-polt 3-metal process. Measurement environment are input voltage 3.7V, output voltage 1.2V and load current range from 10uA to 500mA. And measurement result show that the peak efficiency is 86% and ripple voltage is less 32mV.

• Journal of applied mathematics & informatics
• /
• v.27 no.1_2
• /
• pp.123-137
• /
• 2009

A multiobjective variational problem involving higher order derivatives is considered and Fritz-John and Karush-Kuhn-Tucker type optimality conditions for this problem are derived. As an application of Karush-Kuhn-Tucker optimality conditions, Wolfe type dual to this variational problem is constructed and various duality results are validated under generalized invexity. Some special cases are mentioned and it is also pointed out that our results can be considered as a dynamic generalization of the already existing results in nonlinear programming.

• Journal of the Korean Operations Research and Management Science Society
• /
• v.13 no.2
• /
• pp.9-17
• /
• 1988

This vehicle scheduling problem with multiple depots and multiple vehicle types (VMM) is to determine the optimal vehicle routes to minimize the total travel costs. The object of this paper is to develope an exact algorithm for the VMM. In this paper the VMM is transformed into a mathematical model of the vehicle problem with multiple depots. Then an efficient branch and bound algorithm is developed to obtain an exact solution for this model. In order to enhance the efficiency, this algorithm emphasizes the follows; First, a heuristic algorithm is developed to get a good initial upper bound. Second, an primal-dual approach is used to solve subproblems which are called the quasi-assignment problem, formed by branching strategy is presented to reduce the number of the candidate subproblems.

• Structural Engineering and Mechanics
• /
• v.5 no.5
• /
• pp.491-505
• /
• 1997

This paper deals with the formulation and solution of a wide class of structures, in the presence of both geometric and material nonlinearities, as a particular mathematical programming problem. We first present key ideas for the nonholonomic (path dependent) rate formulation for a suitably discretized structural model before we develop its computationally advantageous stepwise holonomic (path independent) counterpart. A feature of the final mathematical programming problem, known as a nonlinear complementarity problem, is that the governing relations exhibit symmetry as a result of the introduction of so-called nonlinear "residuals". One advantage of this form is that it facilitates application of a particular iterative algorithm, in essence a predictor-corrector method, for the solution process. As an illustrative example, we specifically consider the simplest case of plane trusses and detail in particular the general methodology for establishing the static-kinematic relations in a dual format. Extension to other skeletal structures is conceptually transparent. Some numerical examples are presented to illustrate applicability of the procedure.

• KIPS Transactions on Computer and Communication Systems
• /
• v.10 no.8
• /
• pp.215-222
• /
• 2021

In order to alleviate the limited resource problem and interference problem in cellular networks, the dual connectivity technology has been introduced with the cooperation of small cell base stations. In this paper, we design a new efficient and fair resource allocation scheme for the dual connectivity technology. Based on two different bargaining solutions - Generalizing Tempered Aspiration bargaining solution and Gupta and Livne bargaining solution, we develop a two-stage radio resource allocation method. At the first stage, radio resource is divided into two groups, such as real-time and non-real-time data services, by using the Generalizing Tempered Aspiration bargaining solution. At the second stage, the minimum request processing speeds for users in both groups are guaranteed by using the Gupta and Livne bargaining solution. These two-step approach can allocate the 5G radio resource sequentially while maximizing the network system performance. Finally, the performance evaluation confirms that the proposed scheme can get a better performance than other existing protocols in terms of overall system throughput, fairness, and communication failure rate according to an increase in service requests.

• Korean Journal of Culture and Social Issue
• /
• v.26 no.4
• /
• pp.551-578
• /
• 2020

The purpose of this study is to identify the moderating effect of each of the three stress response (problem-centered treatment, pursuit of social support, and positive thinking) in the effect of work-family conflict between dual-earner couples on marriage satisfaction. To this end, 369 married couples (369 wives, 369 husbands) who raise children under the age of 6 were surveyed on stress coping (problem-centered, social support, and positive thinking), work-family conflicts, and marriage satisfaction. Based on the Actor-Partner Interdependence Model (APIM), the collected data verified six research models by distinguishing the moderating effects of each of the three coping strategies from the direction of Work to Family conflict and Family to Work conflict. Interaction graphs were also presented to determine the pattern of significant buffering effects. As a result, first of all, the problem-oriented strategy of the wife buffer the negative impact of the husband's WFC on the husband's own and wife's marriage satisfaction. It was also found that problem-oriented strategy that husband himself uses to buffer the negative impact of the husband's FWC on his wife's marriage satisfaction. Second, the pursuit of social support confirmed that the negative effects of the husband's WFC on the husband's marriage satisfaction were mitigated by the pursuit of social support used by his wife. Third, in the case of positive thinking, the effect of the positive thinking on the husband's WFC on the marriage satisfaction of the husband and wife was shown, and the positive response effect of the wife's FWC conflict was also shown. Finally, based on the results of this study, the discussion and implications of the study were presented.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.6 no.10
• /
• pp.2473-2492
• /
• 2012

Energy conservation is a vital issue in wireless sensor networks. Recently, employing mobile sinks for data gathering become a pervasive trend to deal with this problem. The sink can follow stochastic or pre-defined paths; however the controlled mobility pattern nowadays is taken more into consideration. In this method, the sink moves across the network autonomously and changes its position based on the energy factors. Although the sink mobility would reduce nodes' energy consumption and enhance the network lifetime, the overhead caused by topological changes could waste unnecessary power through the sensor field. In this paper, we proposed EEDARS, an energy-efficient dual-sink algorithm with role switching mechanism which utilizes both static and mobile sinks. The static sink is engaged to avoid any periodic flooding for sink localization, while the mobile sink adaptively moves towards the event region for data collection. Furthermore, a role switching mechanism is applied to the protocol in order to send the nearest sink to the recent event area, hence shorten the path. This algorithm could be employed in event-driven and multi-hop scenarios. Analytical model and extensive simulation results for EEDARS demonstrate a significant improvement on the network metrics especially the lifetime, the load and the end-to-end delay.

• The Transactions of The Korean Institute of Electrical Engineers
• /
• v.67 no.4
• /
• pp.531-537
• /
• 2018

This paper propose a bidirectional dc-to-dc converter employing dual inductor for current ripple reduction. Conventional bidirectional dc-to-dc converter uses a single inductor for two different modes; boost and buck; therefore it is difficult to satisfy the optimized inductance value for each mode. To improve this problem, the proposed converter adds two switches, a diode, and one inductor. By proper switching of the additional switch, the proposed converter operates with a inductor in boost mode, but it works with dual inductor in buck mode. Hence in both modes the proposed bidirectional converter can be operated with optimized inductance values. Most of all the optimized inductance in buck mode can reduce the current ripple and its effective value(rms), which are directly related to the temperature increase resulted in short lifetime of battery. To verify the validity of the proposed approach, we first analyzes the operation of the proposed converter theoretically, and implement computer-aided simulations and experiments using a prototype.