• Journal of the Semiconductor & Display Technology
• /
• v.16 no.4
• /
• pp.79-85
• /
• 2017

Understanding of human emotion has a high importance in interaction between human and machine communications systems. The most expressive and valuable way to extract and recognize the human's emotion is by facial expression analysis. This paper presents and implements an automatic extraction and recognition scheme of facial expression and emotion through still image. This method has three main steps to recognize the facial emotion: (1) Detection of facial areas with skin-color method and feature maps, (2) Creation of the Bezier curve on eyemap and mouthmap, and (3) Classification and distinguish the emotion of characteristic with Hausdorff distance. To estimate the performance of the implemented system, we evaluate a success-ratio with emotional face image database, which is commonly used in the field of facial analysis. The experimental result shows average 76.1% of success to classify and distinguish the facial expression and emotion.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2019.06a
• /
• pp.252-254
• /
• 2019

본 논문은 합성곱 기반의 Single Shot Multibox Detector(SSD) [1] 의 구조를 이용하여 다양한 스케일의 얼굴들을 잘 검출하도록 하였다. 얼굴 검출은 물체 검출과는 다르게 얼굴의 높이와 너비의 비율이 다소 일정하고 크기가 작은 경우가 많은데, 이에 맞게 얼굴 검출이 용이하도록 anchor의 스케일, 비율, 크기를 변경하였다. 특징점 추출 네트워크는 깊은 네트워크의 최적화를 용이하게 하는 skip connection을 이용한 ResNet-50 [2] 기반을 사용하였다. 다양한 크기, 조명, 환경, 각도의 얼굴들을 포함하는 영상들로 이뤄진 Wider Face[3] 데이터 셋의 easy validation set으로 실험한 결과 0.782과 hard validation set에서 0.611의 average precision을 보였다.

• Research in Community and Public Health Nursing
• /
• v.14 no.3
• /
• pp.375-384
• /
• 2003

The purpose of this study is to identify the amount of social support and quality of life perceived by elderly people in rural areas and to examine the relationship between social support and quality of life in an effort to provide the basic data for effective nursing intervention to enhance quality of life among elderly people. Data were collected from a sample of 220 elderly people aged over 60 from a stratified probability method from one rural area in South Jeolla province from Aug. 15 to Sep. 15, 2002. For data collection, pre-educated researchers visited the participants' homes and surveyed them face-to-face. The statistical analysis was performed using the SAS program along with t-test, ANOVA, and Pearson Correlation. The findings were as follows: 1. The most socially supportive people they answered were Spouses (57.28%), followed by children, friends/neighbors, in descending order. In terms of satisfaction about social support, the subjects felt the most satisfaction from spouses, followed by friends/neighbors, children, in descending order. 2. The participants' perceived social support was 26.0 and the highest score was 108.0 out of 125. Also, the average score and average grade point was 81.7 and 3.35, respectively. A comparison of the average grade points per items within sub-areas of social support revealed financial aid(3.56), informative support(3.34), emotional support (3.27), and evaluation support(3.22). 3. Their perceived social support had statistically significant differences in terms of how much they earned (F=18.56, p= .00001) and whether they had family members living together (F=2.68, p= .0512), quality of life had statistically significant differences in terms of how much they earned (F=35.34, p= 0.00001). 4. The quality of life they perceived was the lowest score 47.0, the highest score 196.0, average score 145.7, and average grade points 3.28. A comparison of the average grade points per items within sub-areas of quality of life revealed neighbor relationships(3.97), self-esteem(3.49), family relationships(3.35), economic conditions 3.12), physical health (2.98), and psychological health(2.74). 5. The relationship between the amount of the subjects' social support and quality of life was significantly correlated (r=.696, p< .001). The findings revealed that social support for elderly people in rural areas was a greatly effective factor on their quality, of life. Also, it was shown that the larger the social support for the elderly, the greater the quality of life for them. Therefore, it is necessary to foster geriatric nurse specialists and develop nursing intervention programs connected with health care and social wellbeing in order to enhance the quality of life of elderly people in rural areas. Also, it is necessary to develop effective models for community and its applications, which will playa leading role for elderly people.

• Maxillofacial Plastic and Reconstructive Surgery
• /
• v.31 no.2
• /
• pp.149-157
• /
• 2009

Purpose: This study was to evaluate the validity of superimposition range at facial images constructed with 3-dimensional (3D) surface laser scanning system. Materials and methods: For the present study, thirty adults, who had no severe skeletal discrepancy, were selected and scanned twice by a 3D laser scanner (VIVID 910, Minolta, Tokyo, Japan) with 12 markers placed on the face. Then, two 3D facial images (T1-baseline, T2-30 minutes later) were reconstructed respectably and superimposed in several manners with $RapidForm^{TM}2006$ (Inus, Seoul, Korea) software program. The distances between markers at the same place of face were measured in superimposed 3D facial images and measurement were done all the 12 makers respectably. Results: The average linear distances between the markers at the same place in the superimposed image constructed by upper 2/3 of the face was $0.92{\pm}0.23\;mm$, in the superimposed image constructed by upper 1/2 of the face was $0.98{\pm}0.26\;mm$, in the superimposed image constructed by upper 1/3 of the face and nose area was $0.99{\pm}0.24\;mm$, in the superimposed image constructed by upper 1/3 of the face was $1.41{\pm}0.48\;mm$, and in the superimposed image constructed by whole face was $0.83{\pm}0.13\;mm$. There were no statistically significant differences in the liner distances of the makers placed on the area included in superimposition range used for partial registration methods but there were significant differences in the linear distances of the markers placed on the areas not included in superimposition range between whole registration method and partial registration methods used in this study. Conclusion: The results of the present study suggest that the validity of superimposition is decreased as superimposition range is reduced in the superimposition of 3D images constructed with 3D laser scanner for the same subject.

• Journal of the Korea Convergence Society
• /
• v.9 no.8
• /
• pp.1-8
• /
• 2018

In the field of biometric security such as face and iris recognition, it is essential to extract facial features such as eyes and lips. In this paper, we have studied a method of detecting eye and lip region in face image using faster R-CNN. The faster R-CNN is an object detection method using deep running and is well known to have superior performance compared to the conventional feature-based method. In this paper, feature maps are extracted by applying convolution, linear rectification process, and max pooling process to facial images in order. The RPN(region proposal network) is learned using the feature map to detect the region proposal. Then, eye and lip detector are learned by using the region proposal and feature map. In order to examine the performance of the proposed method, we experimented with 800 face images of Korean men and women. We used 480 images for the learning phase and 320 images for the test one. Computer simulation showed that the average precision of eye and lip region detection for 50 epoch cases is 97.7% and 91.0%, respectively.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.41 no.6
• /
• pp.121-126
• /
• 2004

An efficient multi face detection technique is proposed based on hierarchical context-free attention operators in which multiple faces are efficiently detected from a noisy and complex background. A noise-tolerant generalized symmetry transform (NTSGT) is applied hierarchically, as a context free attention operator, to the input pyramidal image for the high speed global location of the regions of face candidates (ROFCs) with a single mask. For the face verification, local NTGST is applied within each ROFC to confirm the existence of the detailed facial features. First, by globally applying NTGST which introduces the average pyramid method and focusing to the input image with complex background, ROFCs with recognizable resolution are detected robustly. Morphological operations are applied only to the each detected ROFCs to emphasize the facial features like eyes and lips. Then, eyes are detected by locally appling NTGST to the ROFCs and only faces are detected by verifying the existence of the geometrical features of the faces relatively to the location of eyes. The experimental results show that the proposed method can efficiently detect multiple faces from a noisy or complex background with 93.5% detection rate.

• The Journal of Internal Korean Medicine
• /
• v.31 no.1
• /
• pp.118-127
• /
• 2010

Objective : Hyperhidrosis is the secretion of sweat in amounts greater than physiologically needed for thermoregulation. Ryodoraku is a physiological function test using electric current, and is closely related to skin sympathetic tone. The aim of this study was to investigate the characteristics of Ryodoraku and association of Ryodoraku with hyperhidrosis. Methods : For this study, we conducted Ryodoraku on 56 people, comprising 33 with palmar-plantar hyperhidrosis and 23 with head-face hyperhidrosis, who visited the 5th Internal Department of the Oriental Medicine Hospital of Kyung Hee University during the period from March 1, 2008 to May 31, 2009. Ryodoraku scores were evaluated and compared by 7 indices(upper, lower, left, right, yin, yang, total score). Results : Ryodoraku scores were significantly higher in hyperhidrosis patients than the normal group. Ryodoraku scores of palmar-plantar hyperhidrosis patients were significantly higher than of head-face hyperhidrosis patients and the normal group. The difference of Ryodoraku indices among palmar-plantar hyperhidrosis patients were insignificant. In head-face hyperhidrosis patients, the upper index was significant higher than the lower index. Conclusions : In hyperhidrosis patients, all indices were significant higher than lower. The upper/lower ratio was significant higher in head-face hyperhidrosis patients.

• The Journal of Advanced Prosthodontics
• /
• v.6 no.6
• /
• pp.483-490
• /
• 2014

PURPOSE. The purpose of this study was to determine the average facial proportions and mandibular movement capacity of 316 first-year dental students who carefully recorded them on each other. MATERIALS AND METHODS. This early exacting clinical experience was closely supervised by the authors in Columbus, Ohio during 1969-70. Five vertical and six horizontal distances were measured on each subject's face. An ala-tragus line and an occlusal line were drawn on the left side of the face to determine if these two lines were parallel. Measurements of mandibular movements involved maximum normal and hinge opening at the incisors and maximum amounts of right, left lateral and protrusive excursions of the mandible. RESULTS. The ala width and distance between the tips of upper right and left canine cusps averaged (35.2 mm and 34.8 mm) but with very large individual variations. The distance between ala to occlusal plane lines was 29.9 mm at the tragus and 31.3 mm near the ala. The angle between orbitale and ala-tragus averaged 13.6 degrees. CONCLUSION. The upper lip length was the most variable and the distance between the pupils was the most stable of the eleven facial measurements. The ala-tragus line and the occlusal plane lines were for all practical purposes parallel. Maximum jaw opening averaged 51.2 mm which was 3.0 times larger than maximal hinge opening of 17.2 mm. The maximum right plus left side jaw excursions (9.2 and 9.4 mm) totaled 18.6 mm, 2.3 times more than the 8.0 mm mean maximum forward protrusion.

• Journal of Korea Multimedia Society
• /
• v.10 no.5
• /
• pp.594-603
• /
• 2007

This paper proposes a watermarking scheme for 3D CAD design drawing. In the proposed scheme, we embed binary watermarks into line, arc, and polygon face components that are the basic component of 3D CAD design drawing. The embedding target component can be selected randomly among three components or by the component distribution in drawing. In line components, a watermark bit is embedded into the ratio of the length of a target line and an average length of lines that are connected into a target line. Furthermore, a watermark bit is embedded into a curvature radius on the basis of a center point in a arc component and also is embedded into a ratio of two sides in a polygonal face component. Experimental results verified that the proposed watermarking has the robustness against Format conversion, rotation translation, scaling, cropping, and layer cutting and also SNR of watermarked component is about 39.89-42.50 dB.

• Archives of Craniofacial Surgery
• /
• v.20 no.4
• /
• pp.223-227
• /
• 2019

Background: Neurosensory changes are frequently observed in the patients with mid-face fractures, and these symptoms are often caused by infraorbital nerve (ION) damage. Although ION damage is a relatively common phenomenon, there are no established and objective methods to evaluate it. The aim of this study was to test whether trigeminal somatosensory evoked potential (TSEP) could be used as a prognostic predictor of ION damage and TSEP testing was an objective method to evaluate ION injury. Methods: In this prospective TSEP study, 48 patients with unilateral mid-face fracture (only unilateral blow out fracture and unilateral zygomaticomaxillary fracture were included) and potential ION damages were enrolled. Both sides of the face were examined with TSEP and the non-traumatized side of the face was used as control. We calculated the latency difference between the affected and the unaffected sides. Results: Twenty-four patients recovered within 3 months, and 21 patients took more than 3 months to recover. The average latency difference between the affected side and unaffected side was 1.4 and 4.1 ms for the group that recovered within 3 months and the group that recovered after 3 months, respectively. Conclusion: Patients who suffered ION damage showed prolonged latency when examined using the TSEP test. TSEP is an effective tool for evaluation of nerve injury and predicting the recovery of patients with ION damage.