• Proceedings of the IEEK Conference
• /
• 2003.07e
• /
• pp.2000-2003
• /
• 2003

This paper presents a three-dimensional (3D) head pose estimation algorithm for robust face recognition. Given a 3D input image, we automatically extract several important 3D facial feature points based on the facial geometry. To estimate 3D head pose accurately, we propose an Error Compensated-SVD (EC-SVD) algorithm. We estimate the initial 3D head pose of an input image using Singular Value Decomposition (SVD) method, and then perform a Pose refinement procedure in the normalized face space to compensate for the error for each axis. Experimental results show that the proposed method is capable of estimating pose accurately, therefore suitable for 3D face recognition.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2009.01a
• /
• pp.703-708
• /
• 2009

The present paper describes the application of an improved impression transfer vector method (Sakurai et al., 2007) to transform the three basic dimensions (Evaluation, Activity, and Potency) of higher-order impression. First, a set of shapes and surface textures of faces was represented by multi-dimensional vectors. Second, the variation among faces was coded in reduced parameters derived by applying principal component analysis. Third, a facial attribute along a given impression dimension was analyzed to select discriminative parameters from among principal components with higher sensitivity to impressions, and obtain an impression transfer vector. Finally, the parametric coordinates were changed by adding or subtracting the impression transfer vector and the image was manipulated so that its facial appearance clearly exhibits the transformed impression. A psychological rating experiment confirmed that the impression transfer vector modulated three dimensions of higher-order impression. We discussed the versatility of the impression transfer vector method.

• Proceedings of the IEEK Conference
• /
• 2007.07a
• /
• pp.285-286
• /
• 2007

This paper presents a method to detect eyes in the facial image using Zernike moments and SVM After detecting eye candidate regions from the facial image, Zernike moments are computed on those regions with moving a $15{\times}15$ window. Then, SVM that uses Zernike moments as an input vector detects eyes. In the experimental results, the proposed method shows the eye detection rate of about 90%.

• Journal of the Korean Institute of Telematics and Electronics B
• /
• v.33B no.12
• /
• pp.71-83
• /
• 1996

This paper describes facial component detection and skew correction algorithm for face recognition. We use a priori knowledge and models about isolated regions to detect eye location from the face image captured in natural office environments. The relations between human face components are represented by several rules. We adopt an edge detection algorithm using sobel mask and 8-connected labelling algorith using array pointers. A labeled image has many isolated components. initially, the eye size rules are used. Eye size rules are not affected much by irregular input image conditions. Eye size rules size, and limited in the ratio between gorizontal and vertical sizes. By the eye size rule, 2 ~ 16 candidate eye components can be detected. Next, candidate eye parirs are verified by the information of location and shape, and one eye pair location is decided using face models about eye and eyebrow. Once we extract eye regions, we connect the center points of the two eyes and calculate the angle between them. Then we rotate the face to compensate for the angle so that the two eyes on a horizontal line. We tested 120 input images form 40 people, and achieved 91.7% success rate using eye size rules and face model. The main reasons of the 8.3% failure are due to components adjacent to eyes such as eyebrows. To detect facial components from the failed images, we are developing a mouth region processing module.

• The Research Journal of the Costume Culture
• /
• v.13 no.5 s.58
• /
• pp.712-728
• /
• 2005

This study consists of the stimuli of a female model in her twenties with twenty-three different facial make-up and survey on the differences of them for the development of make-up color system, based on the color-sense on the Korean's skin-tone and make-up color, to enforce the efficiency of beauty education. The result of this study and the suggestion is as followed. Firstly, Familiarity, Intelligence, Fitness, Charm, Tradition and Youth were came out as the result of factor analysis of make-up color image perception. Secondly, the stimulus of bare face was evaluated as more familiar and intelligent than the one with image make-up but perceived as unhealthy and not untraditional. Thirdly, skin tone had a big impact on both in lip color that's been applied in monotonous make-up and in image make-up that had been applied in contrastive make-up. Through these results, it is confirmed that the skin tone and make-up colors were influential variables in the research on facial image perception and preference against a female model in her 20s, and also the image test and preference can be changed according to the color contrasts. This research will be used as a basic tool for the development of make-up color applying system with image perception of statics of population variables and preference research. Also it aims to suggest the alternatives to perform the present collage make-up education for more systematic and organized education.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.17 no.3
• /
• pp.881-895
• /
• 2023

Modern image inpainting techniques based on deep learning have achieved remarkable performance, and more and more people are working on repairing more complex and larger missing areas, although this is still challenging, especially for facial image inpainting. For a face image with a huge missing area, there are very few valid pixels available; however, people have an ability to imagine the complete picture in their mind according to their subjective will. It is important to simulate this capability while maintaining the identity features of the face as much as possible. To achieve this goal, we propose a three-stage network model, which we refer to as the identity and structure feature refinement network (ISFRNet). ISFRNet is based on 1) a pre-trained pSp-styleGAN model that generates an extremely realistic face image with rich structural features; 2) a shallow structured network with a small receptive field; and 3) a modified U-net with two encoders and a decoder, which has a large receptive field. We choose structural similarity index (SSIM), peak signal-to-noise ratio (PSNR), L1 Loss and learned perceptual image patch similarity (LPIPS) to evaluate our model. When the missing region is 20%-40%, the above four metric scores of our model are 28.12, 0.942, 0.015 and 0.090, respectively. When the lost area is between 40% and 60%, the metric scores are 23.31, 0.840, 0.053 and 0.177, respectively. Our inpainting network not only guarantees excellent face identity feature recovery but also exhibits state-of-the-art performance compared to other multi-stage refinement models.

• Journal of KIISE:Software and Applications
• /
• v.27 no.2
• /
• pp.193-200
• /
• 2000

The extraction of face regions is required for Head Gesture Interface which is a natural user interface. Recently, many researchers are interested in using color information to detect face regions in image sequences. Two most widely used color models, HSI color model and YIQ color model, were selected for this study. Actually H-component of HSI and I-component of YIQ are used in this research. Given the difference in the color component, this study was aimed to compare the performance of face region detection between the two models. First, we search the optimum range of facial color for each color component, examining the detection accuracy of facial color regions for variant threshold range about facial color. And then, we compare the accuracy of the face box for both color models by using optimal facial color and motion information. As a result, a range of $0^{\circ}{\sim}14^{\circ}$ in the H-component and a range of $-22^{\circ}{\sim}-2^{\circ}$ in the I-component appeared to be the most optimum range for extracting face regions. When the optimal facial color range is used, I-component is better than H-component by about 10% in accuracy to extract face regions. While optimal facial color and motion information are both used, I-component is also better by about 3% in accuracy to extract face regions.

• Journal of the Institute of Convergence Signal Processing
• /
• v.2 no.2
• /
• pp.47-54
• /
• 2001

The main purpose of this study is to propose the algorithm about the extraction of the facial feature. To achieve the above goal, first of all, this study produces binary image for input color image. It calculates area after pixel labeling by variant block-units. Secondly, by contour following, circumference have been calculated. So the proper degree of resemblance about area, circumference, the proper degree of a circle and shape have been calculated using the value of area and circumference. And Third, the algorithm about the methods of extracting parameters which are about the feature of eyes, nose, and mouse using the proper degree of resemblance, general structures and characteristics(symmetrical distance) in face have been accomplished. And then the feature parameters of the front face have been extracted. In this study, twelve facial feature parameters have been extracted by 297 test images taken from 100 people, and 92.93 % of the extracting rate has been shown.

• The Journal of Korea Institute of Information, Electronics, and Communication Technology
• /
• v.2 no.3
• /
• pp.79-86
• /
• 2009

This study proposed a method of detecting the facial area by calculating Euclidian distances among skin color elements and extracting the characteristics of the face. The proposed algorithm is composed of light calibration and face detection. The light calibration process performs calibration for the change of light. The face detection process extracts the area of skin color by calculating Euclidian distances to the input images using as characteristic vectors color and chroma in 20 skin color sample images. From the extracted facial area candidate, the eyes were detected in space C of color model CMY, and the mouth was detected in space Q of color model YIQ. From the extracted facial area candidate, the facial area was detected based on the knowledge of an ordinary face. When an experiment was conducted with 40 color images of face as input images, the method showed a face detection rate of 100%.

• Journal of Internet Computing and Services
• /
• v.14 no.6
• /
• pp.117-124
• /
• 2013

In this paper, we present a new method which efficiently estimates a face direction from a sequences of input video images in real time fashion. For this work, the proposed method performs detecting the facial region and major facial features such as both eyes, nose and mouth by using the Haar-like feature, which is relatively not sensitive against light variation, from the detected facial area. Then, it becomes able to track the feature points from every frame using optical flow in real time fashion, and determine the direction of the face based on the feature points tracked. Further, in order to prevent the erroneously recognizing the false positions of the facial features when if the coordinates of the features are lost during the tracking by using optical flow, the proposed method determines the validity of locations of the facial features using the template matching of detected facial features in real time. Depending on the correlation rate of re-considering the detection of the features by the template matching, the face direction estimation process is divided into detecting the facial features again or tracking features while determining the direction of the face. The template matching initially saves the location information of 4 facial features such as the left and right eye, the end of nose and mouse in facial feature detection phase and reevaluated these information when the similarity measure between the stored information and the traced facial information by optical flow is exceed a certain level of threshold by detecting the new facial features from the input image. The proposed approach automatically combines the phase of detecting facial features and the phase of tracking features reciprocally and enables to estimate face pose stably in a real-time fashion. From the experiment, we can prove that the proposed method efficiently estimates face direction.