• 제어로봇시스템학회:학술대회논문집
• /
• 2004.08a
• /
• pp.1281-1285
• /
• 2004

There are numerous emotions in the human world. Human expresses and recognizes their emotion using various channels. The example is an eye, nose and mouse. Particularly, in the emotion recognition from facial expression they can perform the very flexible and robust emotion recognition because of utilization of various channels. Hybrid-feature extraction algorithm is based on this human process. It uses the geometrical feature extraction and the color distributed histogram. And then, through the independently parallel learning of the neural-network, input emotion is classified. Also, for the natural classification of the emotion, advancing two-dimensional emotion space is introduced and used in this paper. Advancing twodimensional emotion space performs a flexible and smooth classification of emotion.

• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 2003.09a
• /
• pp.521-524
• /
• 2003

In this paper, we propose a design process of 'personalized' classification with soft computing techniques. Based on human's thinking way, a construction methodology for personalized classifier is mentioned. Here, two fuzzy similarity measures and ensemble of classifiers are effectively used. As one of the possible applications, facial expression recognition problem is discussed. The numerical result shows that the proposed method is very useful for on-line learning, reusability of previous knowledge and so on.

• Journal of the Society of Cosmetic Scientists of Korea
• /
• v.15 no.1
• /
• pp.63-70
• /
• 1989

• Therapeutic Science for Rehabilitation
• /
• v.11 no.1
• /
• pp.69-85
• /
• 2022

Objective : This study aimed to develop a Comprehensive Oro-Facial Function Scale (COFFS) that can evaluate oro-facial function in patients with dysphagia. Methods : To verify the item composition and reliability of the COFFS, preliminary items were collected by selecting and analyzing four previous studies, and the Content Validity Ratio (CVR) was derived through a second survey of experts. Cronbach's 𝛼 was calculated for the internal validity of the evaluation items, and the test-retest reliability and inter-rater reliability were calculated using the internal classification coefficients (ICC). Results : The content validity ratio of all items was 0.67; in the case of Cronbach's 𝛼 value for each domain, 0.849 for communication domain, -0.224 for the oro-facial structure and shape, 0.831 for the ability to perform orofacial movements, and 0.946 for mastication and swallowing function. The test-retest reliability was 0.974 and the inter-rater reliability was 0.937, showing high reliability. Conclusion : In this study, the evaluation tool of COFFS was finally selected from 34 items in four areas and developed on a 3-5 point scale according to the evaluation items. In future studies, additional research is needed to prove its validity through correlation with other evaluation tools that measure oro-facial function.

• Smart Media Journal
• /
• v.13 no.3
• /
• pp.9-17
• /
• 2024

Due to the recent pandemic and the development of ICT technology, the use of non-face-to-face and unmanned systems is expanding, and it is very important to understand emotions in communication in non-face-to-face situations. As emotion recognition methods for various facial expressions are required to understand emotions, artificial intelligence-based research is being conducted to improve facial expression emotion recognition in image data. However, existing research on facial expression emotion recognition requires high computing power and a lot of learning time because it utilizes a large amount of data to improve accuracy. To improve these limitations, this paper proposes a method of recognizing facial expressions using age and gender, which are image meta information, as a method of recognizing facial expressions with even a small amount of data. For facial expression emotion recognition, a face was detected using the Yolo Face model from the original image data, and age and gender were classified through the VGG model based on image meta information, and then seven emotions were recognized using the EfficientNet model. The accuracy of the proposed data classification learning model was higher as a result of comparing the meta-information-based data classification model with the model trained with all data.

• Proceedings of the Korean Society of Computer Information Conference
• /
• 2024.01a
• /
• pp.343-345
• /
• 2024

얼굴 연령 분류 기법은 신원 확인 시스템 고도화, 유동 인구 통계 자동화 시스템 구축, 연령 제한 콘텐츠 관리 시스템 고도화 등 다양한 분야에 적용할 수 있는 확장 가능성을 가진다. 넓은 확장 가능성을 가지는 만큼 적용된 시스템의 안정성을 위해서는 얼굴 연령 분류 기법의 높은 정확도는 필수적이다. 따라서, 본 논문에서는 Vision Transformer(ViT) 기반 분류 알고리즘의 얼굴 연령 분류 성능을 비교 분석한다. ViT 기반분류 알고리즘으로는 최근 널리 사용되고 있는 ViT, Swin Transformer(ST), Neighborhood Attention Transformer(NAT) 세 가지로 선정하였으며, ViT의 얼굴 연령 분류 정확도 65.19%의 성능을 확인하였다.

• Maxillofacial Plastic and Reconstructive Surgery
• /
• v.36 no.6
• /
• pp.266-272
• /
• 2014

Purpose: This study was conducted in order to investigate the therapeutic effect of closed reduction according to a classification in patients with nasal bone fracture. Methods: The study was conducted retrospectively on 186 patients with a mean age of 38 years (range: 7 to 80 years). All patients were diagnosed by clinical and radiologic examination, and then classified according to Hwang's classification by computed tomography. The patients were further classified by their age, gender, causes of fracture, operation timing after fracture, concurrent facial bone fracture, and complications. All patients underwent the same reduction and treatment protocol and were then followed up regularly for at least three months. Results: The cause of the fracture was slip down, and the highest prevalence was shown in the 20s. The mean operation timing after fracture was 4.1 days (range: 1 to 14 days), and it tended to be longer in the case of defected septal bone or more severe fracture. The most common concurrent facial bone fracture was orbital blow-out fracture, and zygomaticomaxillary complex and maxillary fracture occured frequently. The largest number of complications occurred in class III and IIBs patients, and the main complication was postoperative pain. Conclusion: Results of nasal bone closed reduction on the 186 patients showed that serious complications rarely occurred. Closed reduction is generally an effective treatment for nasal bone fracture. However, in the case of severe concurrent septal bone fracture or comminuted fracture with depression, open reduction should be considered. Further study with a larger number of patients and further classification is required.

• Journal of Periodontal and Implant Science
• /
• v.53 no.6
• /
• pp.453-466
• /
• 2023

Purpose: This study aimed to quantify alveolar bone morphology, demonstrate the relationship between tooth angulation and alveolar bone thickness, and introduce a new classification for anterior mandibular teeth related to immediate implant placement (IIP). Methods: Cone-beam computed tomography (CBCT) images of 211 anterior mandibular teeth were analyzed in sagittal slices to measure the thickness of the facial alveolar bone crest (FAB1) and apex (FAB2), and the lingual alveolar bone crest (LAB1) and apex (LAB2). Tooth angulation was classified as 1°-10°, 11°-20°, and >20° according to the tooth's long axis and alveolar bone wall. Spearman correlation coefficients were used to evaluate correlations between the variables. Results: FAB1 and LAB1 were predominantly thin (<1 mm) (84.4% and 73.4%, respectively), with the lateral incisors being thinnest. At the apical level, FAB2 and LAB2 were thick in 99.5% and 99.1% of cases, respectively. Significant differences were documented in FAB2 (P=0.004), LAB1 (P=0.001), and LAB2 (P=0.001) of all mandibular teeth. At all apical levels of the inspected teeth, a significant negative correlation existed between TA and FAB2. Meanwhile, TA showed a significant positive correlation with LAB2 of the lateral incisors and canines. These patterns were then divided into class I (thick facial and lingual alveolar bone), class II (facially inclined teeth) with subtype A (1°-10°) and subtype B (11°-20°), and class III (lingually inclined teeth) with subtype A (1°-10°) and subtype B (11°-20°). Conclusions: Mandibular anterior teeth have predominantly thin facial and lingual crests, making the lingual bone apical thickness crucial for IIP. Although anchorage can be obtained from lingual bone, tooth angulation and tooth types had an impact on IIP planning. Hence, the new classification based on TA and alveolar bone wall may enable rational clinical planning for IIP treatment.

• Maxillofacial Plastic and Reconstructive Surgery
• /
• v.34 no.1
• /
• pp.26-33
• /
• 2012

Purpose: The golden ratio has been used for a long time to objectify and quantify 'beauty'. Dr. Marqurardt claims that the golden ratio can be applied in the maxillofacial field as well. The purpose of this study was to evaluate the diagnostic significance of using a facial 'phi' mask for analyzing Korean faces with characteristics of Class I, II, and III malocclusion. Methods: We studied twenty five Korean celebrities' frontal facial photos (10 males, 15 females) and 90 malocclusion patients' frontal facial photos (30 patients in each malocclusion classification: Class I, Class II, and Class III). Patients who received orthodontic treatment at Samsung Medical Center were selected for this study. After superimposition of the selected facial photo and facial 'phi' mask using Adobe Photoshop CS3, the ratio of the entire facial area, mid facial area, lower facial area and horizontal and vertical lengths were measured. Results: The facial ratio in photos of Korean faces showed larger vertical and horizontal ratios than the facial 'phi' mask with golden ratio, regardless of skeletal malocclusion (entire face: 115%, lower face: 125% larger than the mask). The results of the frontal photos of Class I, II, and III malocclusion patients using facial 'phi' mask showed that the vertical length and frontal face area was more significantly influenced by the area of the lower face than the midface. This means that the lower face has larger proportions in the facial areas. Conclusion: The ratio of facial 'phi' mask is matched with the ideal facial appearance that the contemporary Korean general public is seeking. Thus, the facial 'phi' mask may be a convenient tool for esthetic analysis of Korean faces. Reducing the area of the lower face is esthetically more desirable for almost all Korean people when planning orthognathic surgery.

• Journal of Korean Neurosurgical Society
• /
• v.40 no.6
• /
• pp.445-449
• /
• 2006

Objective : The accurate anatomy of the facial nerve is essential for successful surgical outcome. The purpose of the present study is to know such information on the facial nerve from a series of specimens. Methods : This study is based on cadaveric dissection of 41 Korean fetus and stillborn infant and describes anatomical variations of the peripheral branches of the facial nerve that pose a importance in a number of neruosurgical procedure. Results : The branching patterns were classified into six types according to modified Davis classification : the frequencies of occurrence were : type I, 4.9%; type II, 24.4%; type III, 34.1%; type IV, 19.5%; type V, 12.2%; and type VI, 4.9%. Types II, III and IV together accounted for almost 80% of the specimens. Conclusion : Compared to previous adult and western stillborn fetus cadaveric studies, there was no significant difference in the percentage of the types between the subjects in the present study, similar pattern and anatomic distribution.