• International Journal of Internet, Broadcasting and Communication
• /
• v.13 no.2
• /
• pp.187-194
• /
• 2021

This paper presents a margin-based face liveness detection method with behavioral confirmation to prevent spoofing attacks using deep learning techniques. The proposed method provides a possibility to prevent biometric person authentication systems from replay and printed spoofing attacks. For this work, a set of real face images and fake face images was collected and a face liveness detection model is trained on the constructed dataset. Traditional face liveness detection methods exploit the face image covering only the face regions of the human head image. However, outside of this region of interest (ROI) might include useful features such as phone edges and fingers. The proposed face liveness detection method was experimentally tested on the author's own dataset. Collected databases are trained and experimental results show that the trained model distinguishes real face images and fake images correctly.

• Proceedings of the KIEE Conference
• /
• 2008.10b
• /
• pp.333-334
• /
• 2008

A real-time face tracking is a broad topic, covering a large spectrum of technologies and applications. Briefly face tracking is a kind of tracing technique which follows human face in any directions. It needs some algorithms such as human face detection and motion controller to track face. Moreover, both processing time and calculation time are the most important factors that influence to drive tracking system. In this paper, two algorithms are used to find human face: earn-shift algorithm and face detection algorithm using OpenCV. Fuzzy controller is utilized to move pan-tilt camera system which can move four directions along to x-y axis.

• The Research Journal of the Costume Culture
• /
• v.6 no.4
• /
• pp.13-25
• /
• 1998

This study focuses on the visual effects of various neckline-haristyle combinations on the perception of face image and type. The study employed a model with oval face and examined 35 combinations made up of five necklines and seven hairstyles. Looking at various face images depending upon different necklines, in case of round, V, boat square, and high necklines, long wave hair and medium wave hair produced a feminine image of marked individuality; long straight hair that covers the forehead and medium straight hair gave an image of charm and purity as well as an image of neatness; and long straight hair short cut hair showed an intellectual image. Regarding the perceptual type of face depending upon the different necklines of round, V, square, and high, the long straight hair covering the forehead and medium wave hair had the effect of an optical illusion that made the face look short and round; and short cut hair made the face line look distinct and the face look oval and slender.

• Journal of Institute of Control, Robotics and Systems
• /
• v.10 no.11
• /
• pp.991-999
• /
• 2004

Face robots capable of expressing their emotional status, can be adopted as an efficient tool for friendly communication between the human and the machine. In this paper, we present a face robot actuated with artificial muscle based on dielectric elastomer. By exploiting the properties of polymers, it is possible to actuate the covering skin, eyes as well as provide human-like expressivity without employing complicated mechanisms. The robot is driven by seven types of actuator modules such as eye, eyebrow, eyelid, brow, cheek, jaw and neck module corresponding to movements of facial muscles. Although they are only part of the whole set of facial motions, our approach is sufficient to generate six fundamental facial expressions such as surprise, fear, anger, disgust, sadness, and happiness. Each module communicates with the others via CAN communication protocol fur the desired emotional expressions, the facial motions are generated by combining the motions of each actuator module. A prototype of the robot has been developed and several experiments have been conducted to validate its feasibility.

• Journal of Korea Multimedia Society
• /
• v.15 no.4
• /
• pp.508-515
• /
• 2012

This paper presents a method that estimates face pose for a given face image and synthesizes any posed face images using Active Appearance Model(AAM). The AAM that having been successfully applied to various applications is an example-based learning model and learns the variations of training examples. However, with a single model, it is difficult to handle large pose variations of face images. This paper proposes to build a model covering only a small range of angle for each pose. Then, with a proper model for a given face image, we can achieve accurate pose estimation and synthesis. In case of the model used for pose estimation was not trained with the angle to synthesize, we solve this problem by training the linear relationship between the models in advance. In the experiments on Yale B public face database, we present the accurate pose estimation and pose synthesis results. For our face database having large pose variations, we demonstrate successful frontal pose synthesis results.

• Tunnel and Underground Space
• /
• v.18 no.6
• /
• pp.427-435
• /
• 2008

In this paper, review was made for the excavation method and optimum bench length for unstable tunnel face in case of rock classification type V in order to make the best use of in-situ bearing capacity. 3D FEM analyses were performed to investigate the influences on the tunnel face and adjacent area with regard to the pattern and number of bolts when face bolts were used as a supplementary measure. As a result of this study, full section excavation method with sub-bench is effective in reducing the displacement greatly due to early section closure. Displacement-resistant effects in accordance with the bolting patterns are grid type, zig-zag type and then circular type in order of their effect. And horizontal extrusion displacement of tunnel face reduces as the number of bolts increase. A grid type face bolt covering $1.5m^2$ of tunnel face could secure the face stability in case of full section excavation method with sub-bench.

• Structural Engineering and Mechanics
• /
• v.54 no.5
• /
• pp.955-971
• /
• 2015

In this paper, the influence of the polled direction of piezoelectric materials on the stress distribution is studied under time-harmonic dynamical load (time-harmonic Lamb's problem). The system considered in this study consists of piezoelectric covering layer and piezoelectric half-plane, and the harmonic dynamical load acts on the free face of the covering layer. The investigations are carried out by utilizing the exact equations of motion and relations of the linear theory of electro-elasticity. The plane-strain state is considered. It is assumed that the perfect contact conditions between the covering layer and half-plane are satisfied. The boundary value problems under consideration are solved by employing Fourier exponential transformation techniques with respect to coordinates directed along the interface line. Numerical results on the influence of the polled direction of the piezoelectric materials such as PZT-5A, PZT-5H, PZT-4 and PZT-7A on the normal stresses, shear stresses and electric potential acting on the interface plane are presented and discussed. As a result of the analyses, it is established that the polled directions of the piezoelectric materials play an important role on the values of the studied stresses and electric potential.

• 제어로봇시스템학회:학술대회논문집
• /
• 2004.08a
• /
• pp.1229-1234
• /
• 2004

Face robots capable of expressing their emotional status, can be adopted as an e cient tool for friendly communication between the human and the machine. In this paper, we present a face robot actuated with arti cial muscle based on dielectric elastomer. By exploiting the properties of polymers, it is possible to actuate the covering skin, and provide human-like expressivity without employing complicated mechanisms. The robot is driven by seven types of actuator modules such as eye, eyebrow, eyelid, brow, cheek, jaw and neck module corresponding to movements of facial muscles. Although they are only part of the whole set of facial motions, our approach is su cient to generate six fundamental facial expressions such as surprise, fear, angry, disgust, sadness, and happiness. Each module communicates with the others via CAN communication protocol and according to the desired emotional expressions, the facial motions are generated by combining the motions of each actuator module. A prototype of the robot has been developed and several experiments have been conducted to validate its feasibility.

• Journal of Mechanical Science and Technology
• /
• v.19 no.2
• /
• pp.578-588
• /
• 2005

Face robots capable of expressing their emotional status, can be adopted as an efficient tool for friendly communication between the human and the machine. In this paper, we present a face robot actuated with artificial muscle based on dielectric elastomer. By exploiting the properties of dielectric elastomer, it is possible to actuate the covering skin, eyes as well as provide human-like expressivity without employing complicated mechanisms. The robot is driven by seven actuator modules such eye, eyebrow, eyelid, brow, cheek, jaw and neck module corresponding to movements of facial muscles. Although they are only part of the whole set of facial motions, our approach is sufficient to generate six fundamental facial expressions such as surprise, fear, angry, disgust, sadness, and happiness. In the robot, each module communicates with the others via CAN communication protocol and according to the desired emotional expressions, the facial motions are generated by combining the motions of each actuator module. A prototype of the robot has been developed and several experiments have been conducted to validate its feasibility.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.21 no.7
• /
• pp.285-291
• /
• 2020

In manufacturing, humans are being replaced with robots, but expert skills remain difficult to convert to data, making them difficult to apply to industrial robots. One method is by visual motion recognition, but physical features may be judged differently depending on the image data. This study aimed to improve the accuracy of vision methods for estimating the posture of humans. Three OpenPose vision models were applied: MPII, COCO, and COCO+foot. To identify the effects of face-covering accessories and image preprocessing on the Convolutional Neural Network (CNN) structure, the presence/non-presence of accessories, image size, and filtering were set as the parameters affecting the identification of a human's posture. For each parameter, image data were applied to the three models, and the errors between the actual and predicted values, as well as the percentage correct keypoints (PCK), were calculated. The COCO+foot model showed the lowest sensitivity to all three parameters. A <50% (from 3024×4032 to 1512×2016 pixels) reduction in image size was considered acceptable. Emboss filtering, in combination with MPII, provided the best results (reduced error of <60 pixels).