• ETRI Journal
• /
• v.44 no.2
• /
• pp.286-299
• /
• 2022

Human activity recognition in real time is a challenging task. Recently, a plethora of studies has been proposed using deep learning architectures. The implementation of these architectures requires the high computing power of the machine and a massive database. However, handcrafted features-based machine learning models need less computing power and very accurate where features are effectively extracted. In this study, we propose a handcrafted model based on three-dimensional sequential skeleton data. The human body skeleton movement over a frame is computed through joint positions in a frame. The joints of these skeletal frames are projected into two-dimensional space, forming a "movement polygon." These polygons are further transformed into a one-dimensional space by computing amplitudes at different angles from the centroid of polygons. The feature vector is formed by the sampling of these amplitudes at different angles. The performance of the algorithm is evaluated using a support vector machine on four public datasets: MSR Action3D, Berkeley MHAD, TST Fall Detection, and NTU-RGB+D, and the highest accuracies achieved on these datasets are 94.13%, 93.34%, 95.7%, and 86.8%, respectively. These accuracies are compared with similar state-of-the-art and show superior performance.

• Journal of Information Processing Systems
• /
• v.12 no.3
• /
• pp.381-391
• /
• 2016

The Joint Bayesian (JB) method has been used in most state-of-the-art methods for face verification. However, since the publication of the original JB method in 2012, no improved verification method has been proposed. A lot of studies on face verification have been focused on extracting good features to improve the performance in the challenging Labeled Faces in the Wild (LFW) database. In this paper, we propose an improved version of the JB method, called the two-dimensional Joint Bayesian (2D-JB) method. It is very simple but effective in both the training and test phases. We separated two symmetric terms from the three terms of the JB log likelihood ratio function. Using the two terms as a two-dimensional vector, we learned a decision line to classify same and not-same cases. Our experimental results show that the proposed 2D-JB method significantly outperforms the original JB method by more than 1% in the LFW database.

• Maxillofacial Plastic and Reconstructive Surgery
• /
• v.37
• /
• pp.28.1-28.7
• /
• 2015

Background: The aim of this study was to present three-dimensional (3D) structural characteristics of the mandible in the hemifacial microsomia. The mandible has six distinct functional units, and its architecture is the sum of balanced growth of each functional unit and surrounding matrix. Methods: In order to characterize the mandibular 3D architecture of hemifacial microsomia, we analyzed the mandibular functional units of four hemifacial microsomia patients using the 3D reconstructed computed tomography (CT) images. And we compared the functional unit size between affected and non-affected side. Results: The length of condyle and angle showed significant differences between affected and non-affected sides. However, the length of mandibular body showed insignificant differences. The size differences between affected and non-affected side were observed at the condyle, angle, and body in descending order. Conclusions: This preliminary study suggests that the main etiopathogenic units are condyle and angle in the hemifacial microsomia mandible. Further investigation with the increased number of subjects will be helpful to establish treatment modality by etiopathogenic targeting of hemifacial microsomia.

• Journal of Microbiology and Biotechnology
• /
• v.32 no.2
• /
• pp.238-247
• /
• 2022

Since the skin covers most surfaces of the body, it is susceptible to damage, which can be fatal depending on the degree of injury to the skin because it defends against external attack and protects internal structures. Various types of artificial skin are being studied for transplantation to repair damaged skin, and recently, the production of replaceable skin using three-dimensional (3D) bioprinting technology has also been investigated. In this study, skin tissue was produced using a 3D bioprinter with human skin cell lines and cells extracted from mouse skin, and the printing conditions were optimized. Gelatin was used as a bioink, and fibrinogen and alginate were used for tissue hardening after printing. Printed skin tissue maintained a survival rate of 90% or more when cultured for 14 days. Culture conditions were established using 8 mM calcium chloride treatment and the skin tissue was exposed to air to optimize epidermal cell differentiation. The skin tissue was cultured for 14 days after differentiation induction by this optimized culture method, and immunofluorescent staining was performed using epidermal cell differentiation markers to investigate whether the epidermal cells had differentiated. After differentiation, loricrin, which is normally found in terminally differentiated epidermal cells, was observed in the cells at the tip of the epidermal layer, and cytokeratin 14 was expressed in the lower cells of the epidermis layer. Collectively, this study may provide optimized conditions for bioprinting and keratinization for three-dimensional skin production.

• Geophysics and Geophysical Exploration
• /
• v.5 no.4
• /
• pp.236-249
• /
• 2002

We have extended the three-dimensional (3-D) resistivity imaging algorithm to cover the 3-D resistivity tomography problem, where resistivity data are acquired using electrodes installed in several boreholes as well as at the earth surface. The imaging algorithm consists of the 3-D finite element forward modeling and least-squares inversion scheme, where the ACB (Active Constraint Balancing) is adopted to enhance the resolving power of the inversion. Sensitivity analysis with numerical verifications shows that 3-D resistivity tomography is a very appealing method and can be used to get 3-D attitude of subsurface structures with very high-resolution. Moreover, we could accurately handle the topography effect, which could cause artifacts in the resistivity tomography. In the application of 3-D resistivity tomography to the real field data set acquired at the quarry mine, we could derive a very reasonable and accurate image of the subsurface.

• Journal of Korean Neurosurgical Society
• /
• v.64 no.6
• /
• pp.882-890
• /
• 2021

Objective : The aim of this study to investigate the benefits of patient-based 3-dimensional (3D) cerebral arteriovenous malformation (AVM) models for preoperative surgical planning and education. Methods : Fifteen patients were operated on for AVMs between 2015 and 2019 with patient-based 3D models. Ten patients' preoperative cranial angiogram screenings were evaluated preoperatively or perioperatively via patient-based 3D models. Two patients needed emergent surgical intervention; their models were solely designed based on their AVMs and used during the operation. However, the other patients who underwent elective surgery had the modeling starting from the skull base. These models were used both preoperatively and perioperatively. The benefits of patients arising from treatment with these models were evaluated via patient files and radiological data. Results : Fifteen patients (10 males and five females) between 16 and 66 years underwent surgery. The mean age of the patients was 40.0±14.72. The most frequent symptom patients observed were headaches. Four patients had intracranial bleeding; the symptom of admission was a loss of consciousness. Two patients (13.3%) belonged to Spetzler-Martin (SM) grade I, four (26.7%) belonged to SM grade II, eight (53.3%) belonged to SM grade III, and one (6.7%) belonged to SM grade IV. The mean operation duration was 3.44±0.47 hours. Three patients (20%) developed transient neurologic deficits postoperatively, whereas three other patients died (20%). Conclusion : Several technological innovations have emerged in recent years to reduce undesired outcomes and support the surgical team. For example, 3D models have been employed in various surgical procedures in the last decade. The routine usage of patient-based 3D models will not only support better surgical planning and practice, but it will also be useful in educating assistants and explaining the situation to the patient as well.

• Journal of Welding and Joining
• /
• v.18 no.3
• /
• pp.76-82
• /
• 2000

In this study, three-dimensional heat flow in laser beam welding for deep penetration was analyzed by using F.E.M common code, and then the results were compared with the experimental data. The models for analysis are full penetration welds and are made at three different laser powers (6, 9.9, 4.5 kW) with two different welding speeds (5.8mm/s, 5mm/s). The characteristics of thermal absorption by the workpiece during deep penetration laser welding can be represented by a combination of line heat source through the workpiece and distributed heat source at the top surface due to the plasma plume above the top surface. This gives an insight into the way in which the beam interacts with the material being welded. The analyses performed with the combined heat source models show comparatively good agreement between the experimental and calculated melt temperature isotherm, i.e, the fusion zone boundary. The results are used to explain the "nail head" appearance of fusion zone, which is quite common in laser beam welds.eam welds.

• Proceedings of the Korean Society of Marine Engineers Conference
• /
• 2005.06a
• /
• pp.210-215
• /
• 2005

A three-dimensional numerical simulation is performed to investigate on a low Reynolds number mixed convection in a horizontal rectangular channel with the upper part cooled and the lower part heated uniformly. The three-dimensional governing equations are solved using a finite volume method. For convective term, the central differencing scheme is used and for the pressure correction, the PISO algorithm is used. Solutions are obtained for A=4, Pr=0.72, 10, 909, the Reynolds number ranging from $2.1{\times}10^{-2}$ to $1.2{\times}10^{-1}$, the Rayleigh number is $3.5{\times}10^4$. It is found that vortex roll structures of mixed convection in horizontal rectangular channel can be classified into three roll structures which affected by Prandtl number and Reynolds number.

• The Transactions of The Korean Institute of Electrical Engineers
• /
• v.63 no.1
• /
• pp.162-172
• /
• 2014

A three-dimensional dynamic model for simulating various motions of full vehicle is presented. The model has 16 independent degrees of freedom (DOF) consisting of three kinds of components; a vehicle body of 6 DOF, 4 independent suspensions equipped at every corner of the body, and 4 tire models linked with each suspension. The dynamic equations are represented in six coordinate frames such as world fixed coordinate, vehicle fixed coordinate, and four wheel fixed coordinate frames. Then these lead to the approximated prediction model of vehicle posture. Both lateral and longitudinal dynamics can be computed simultaneously under the conditions of which various inputs including steering command, driving torque, gravity, rolling resistance of tire, aerodynamic resistance, etc. are considered. It is shown through simulations that the proposed 3D model can be useful for precise design and performance analysis of any full vehicle control systems.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.36 no.9C
• /
• pp.573-580
• /
• 2011

WSNs (Wireless Sensor Networks) are becoming more widely used in various fields, and improving localization performance is a crucial and essential issue for sensor network applications. In this paper, we propose a low-complexity localization mechanism for WSNs that operates in 3D (Three-Dimensional) space. The basic idea is to use aerial vehicles or flying objects that are deliberately equipped with the anchor nodes. These anchor nodes periodically broadcast beacon signals containing their current locations, and the unknown nodes receive these signals as soon as they enter the communication range of the anchors. We estimate the locations of the unknown nodes based on the proposed scheme that transforms the 3D problem into 2D computations to reduce the complexity of 3D localization. Simulated results show that our approach is an effective scheme for 3D self-positioning in WSNs.