• Journal of Information Processing Systems
• /
• v.16 no.1
• /
• pp.6-29
• /
• 2020

Biometrics identification using multiple modalities has attracted the attention of many researchers as it produces more robust and trustworthy results than single modality biometrics. In this paper, we present a novel multimodal recognition system that trains a deep learning network to automatically learn features after extracting multiple biometric modalities from a single data source, i.e., facial video clips. Utilizing different modalities, i.e., left ear, left profile face, frontal face, right profile face, and right ear, present in the facial video clips, we train supervised denoising auto-encoders to automatically extract robust and non-redundant features. The automatically learned features are then used to train modality specific sparse classifiers to perform the multimodal recognition. Moreover, the proposed technique has proven robust when some of the above modalities were missing during the testing. The proposed system has three main components that are responsible for detection, which consists of modality specific detectors to automatically detect images of different modalities present in facial video clips; feature selection, which uses supervised denoising sparse auto-encoders network to capture discriminative representations that are robust to the illumination and pose variations; and classification, which consists of a set of modality specific sparse representation classifiers for unimodal recognition, followed by score level fusion of the recognition results of the available modalities. Experiments conducted on the constrained facial video dataset (WVU) and the unconstrained facial video dataset (HONDA/UCSD), resulted in a 99.17% and 97.14% Rank-1 recognition rates, respectively. The multimodal recognition accuracy demonstrates the superiority and robustness of the proposed approach irrespective of the illumination, non-planar movement, and pose variations present in the video clips even in the situation of missing modalities.

• The korean journal of orthodontics
• /
• v.13 no.2
• /
• pp.177-183
• /
• 1983

The soft tissue covering of the face plays an important role in facial esthetics, speech and other physiologic functions. Thus, it is recognised by all clinical orthodontists that success of orthodontic treatment is closely related to the changes in soft tissues of the face. The purpose of this study was to evaluate the changes of bony and soft tissues in prepost treatment of Angle's Class III malocclusion. The sample consisted of 18 males and 37 females, pretreatment age of 9 years to 11 years. For this study 11 landmarks were plotted, 14 linear length, 4 soft tissue thickness and 2 angles were measured. The obtained results were as follows; 1. In the linear measurements of bony and soft tissue changes, A, Is, Ss, Ls and Li were located more anteriorly in both sexes. However Si and B showed more remarkable anterior movement in female. 2. In the comparison of the changes of the soft tissue thickness, Ss and Li in male subjects and Ss in female subjects increased. 3. In the degree of correlation between changes in the soft tissue profile and changes in the skeletal profile, Is: Ls, Il: Li and B: Si in both sexes had significant correlations. However A:Ss had remarkable significant correlation in female. 4. There were significant correlations between change in ${\angle}A$ and change in ${\angle}B$ in all sexes. 5. There were little correlation between changes in distance difference of Is and Ii and Change in distance difference of Ls and Li in all sexes.

• Journal of Korean Academy of Oral and Maxillofacial Radiology
• /
• v.8 no.1
• /
• pp.49-61
• /
• 1978

A study was made to investigate a relationship existing in the dentoskeletal framework and the soft tissue profile around the face, and compare the sexual differences between boys and girls having the normal occlusion in the mixed dentition. The lateral cephalograms were composed of 67 boys aged 10.3 years and 68 girls aged 10.4 years, respectively. By means of the lateral cephalograms, both the hard-and soft tissue structures were simultaneously analyzed, measured and evaluated by introducing the several reference items: S-N plane, palatal plane, mandibular plane, N-A line and A-P line for the dentoskeletal structures and N’-P’line for the soft tissue, and the 21 measuring points for the both structures. The significant findings were as follows: 1. In general the boys showed the larger nasal component dimension than did the girls, but the length and height of nose(N’-Prn and NA-Prn) showed the significant sexual difference among those when evaluated statistically. 2. The lip-thickness was found to be minimal in the region of nasion, greater in the region of pogonion, and much greater in the region of point A in the both sexes, but the only thickness over point A(A-A’) showed the significant sexual difference statistically. 3. The upper and lower lip position were found to be located anteriorly to the esthetic line in the both sexes, but upper lip position showed the significant sexual difference when evaluated statistically. 4. The regions of nose and upper lip had a tendency not to be following the underlying skeletal profile.

• The korean journal of orthodontics
• /
• v.27 no.2
• /
• pp.245-258
• /
• 1997

Cephalometric radiographs, frontal photographs and profile silhouette phogographs of 68 young adult female who were model or were recommended to have esthetic face were used in this study. 7 Students in department of Art of Kyungpook national university and 15 orthodontists estimated profile slides which were made of 3 Profile silhouettes in parallel with FH plane. Profile silhouettes were made of soft tissue profile line of cephalometric radiograph. Only orthodontists estimated frontal photographs. Students and orthodontists score 9 in excellent case, score 7 in good case, score 5 in average case, score 3 in poor case. Correlation analysis between orthodontists' esthetic concept and Artists' esthetic concept, between frontal view esthetics and profile view esthetics which estimated by orthotontists, between profile view esthetics and profile measurements which consisted of measurements of 38 female who were scored above 5 mean score in profile silhouette by orthodontists were done. And the finding in this study indicated the following 1. Correlation between orthodontists' esthetic concept and Artists' esthetic concept in profile silhouette was significant (r=0.67,P=0.0001). 2. Correlation between frontal view esthetics and profile view esthetics which estimated by orthodontist was significant (r=0.26,P=0.0381). 3. Measurements which had significant correlation between profile measurements and profile view esthetics wer Na-Pog, to N', BNV to Pog', BNV/B' -Pog', Ls-Li-Pog', Li-B'-Pog' Z angle(P<0.05). 4 Mean and standard deviation of profile measurements of 38 female were obtained.

• The korean journal of orthodontics
• /
• v.6 no.1
• /
• pp.17-24
• /
• 1976

The primary objective of this study was to define the differences that exist. between different sexes on the dentoskeletal framework and the soft tissue profile around the mouth. For the purpose of this study, cephalometric radiographs were obtained from the centric occlusion with closed lip position, through the research on each 42 males and females aged from 17 to 22 years with normal occlusion and acceptable facial appearence. The results were as follows: 1. Maxillary to mandibular relationships. Among the angles formed by the long axis of the maxillary and mandibular anterior teeth, the maxillary and mandibular anterior alveolar bone, and the lower and upper lips (Fig.2), only the angle formed by the lips was more acute in males than in females. The males have a more rounded profile, and the females have a flatter profile in the lower third of the face. The differences is statistically significant for the angle formed by the lips. The fact that the lips have a difference greater than that of teeth or the alveolar bone indicates that the lip position is not entirely due to tooth and bony support. Possibly the thickness of the lips has an influence. 2. Occlusal plane. The occlusal plane was related to the anterior tooth inclination, anterior alveolar bone profile, and the lip contour, both maxillary and mandibular (Fig.3). Only the angle related to lower lip was statistically significant. The females again had the more obtuse angle, indicating a flatter profile than that of the males. 3.Skeletal planes. The angles formed by the anterior maxillary lips, teeth, and alveolar bone with the Frankfort plane and the angles of the mandibular lips, teeth, and, alveolar bone and the mandibular plane were investigated (Fig.4). Results were similar to those from maxillary to mandibular relationships. The results were statistically significant for the upper lip and the lower lip, only. 4. Esthetics. The facial line and the mandibular plane were compared with the esthetic line. These angles were different for the different sexes, but only the latter was statistically significant. This difference may be due to the profile contour of the nose.

• Archives of Plastic Surgery
• /
• v.40 no.4
• /
• pp.433-439
• /
• 2013

Background Facial beauty depends on the form, proportion, and position of various units of the face. In terms of the frontal view and facial profile, the chin is the most prominent aesthetic element of the lower third of the face. Many methods have been implemented to obtain good proportions of the lower face. In this study, we applied the T-shaped genioplasty method to correcting chin deformities. Methods All of the procedures in 9 cases were performed under general anesthesia. For genioplasty, a horizontal cutting line and 1 or 2 vertical cutting lines were drawn 5 mm below the mental foramen. Osteotomed bone segments of the chin were used for horizontal widening using bone grafts or for horizontal shortening. Likewise, they were used as bone grafts for vertical lengthening or vertical shortening. The bone segments were approximated in the midline and held in place using miniplates. Results The postoperative appearance of the 9 cases showed that the lower third of the face had been naturally changed. At the same time, vertical lengthening or shortening, and horizontal widening or shortening could be implemented during the operation. Satisfactory results were obtained based on reviews of the patients' preoperative and postoperative photographs. The patients were also satisfied with the outcomes. Conclusions Using T-shaped genioplasty, we efficiently adjusted the shape and position of the chin to obtain good proportions of the lower face and change its contour to obtain an aesthetically appealing oval face in accordance with East Asians' aesthetic preferences.

• The Journal of the Korea Contents Association
• /
• v.11 no.11
• /
• pp.91-101
• /
• 2011

This research aims to suggest web-based 3D face modeling system for hairline modification surgery. As public interests in beauty regarding face escalate with era of wide persoanl mobile smart iCT devices, need for medical information system is urgent and increasing demand. This research attempted to build 3D facing modeling library deploying conventional technology and proprietary software available. Implications from the our experiment found that problems and requirement for developing new web based standard. We suggest new system from our experiment and literature review regarding relevant technologies. Main features of our suggested systems is based on studies regarding hair loss treatment such as medical science, beauty studies and information technology. This system processes input images of 2D frontal and profile pictures of face into 3D face modeling with mesh-data. The mesh data is compatible with web standard technology including SVG and Canvas Tag supported natively by HTML5.

• Maxillofacial Plastic and Reconstructive Surgery
• /
• v.22 no.2
• /
• pp.174-183
• /
• 2000

This study was performed to investigate the characteristics of soft tissue profile of the class III malocclusion and to test the yardstick far differential diagnosis between surgical and orthodontic patients. Initial lateral cephalograms of orthodontic group(30 patients) that have acceptable occlusion and profile by orthodontic treatment alone and surgical group(30 patients) that have favorable occlusion and profile by combined surgical-orthodontic treatment were selected in Ajou university hospital. Powell and Burstone II analysis were made on the tracing. Descriptive, comparative, factor, cluster, and discriminant analysis were carried out with computer program. The results were as followings : 1. Patients who received surgery had a more concave profile and a longer lower facial height than patients who received orthodontic treatment alone. 2. Nasolabial angle, ratio of vertical height, and mentolabial sulcus were significantly different at the 5% level. And facial protuberance, upper lip protuberance, mentocervical angle, nasofrontal angle, nasomental angle, mandibular vertical height, angle between cervix and lower face, ratio of mandibular vertical height divided by cervical depth, ratio of vertical height between upper and lower lip, and maxillary protuberance were significantly different at the 1% level. 3. 8 factors were extracted and factor 2, 3, and 8 showed significant differences by factor analysis. 4. Orthodontic group (25) and surgical group (35) were classified by cluster analysis. 5. Discriminant function was D = 0.079Nasomental angle + 0.081Sn-Gn + 3.343Sn-Gn/C-Gn + 1.734Sn-St/St-Me' -26.460, and cutting score was 0, so we can discriminate that orthodontic group has the score above 0, and surgery group below 0. And 91.7% of original grouped cases were correctly classified.

• Journal of the Korean Society for Precision Engineering
• /
• v.1 no.3
• /
• pp.85-94
• /
• 1984

The WILDHABER-NOVIKOV gear, one of the circular arc gears, has the large contact area between the convex and concave profiled mating teeth, moves from one end of the tooth to the other axially making a face contact. Hence it provides a large load capacity than the Involute gear and still satisfying the law of gearing. In order to analyze the gear stress, a photoelastic investigation was carried out. Photo elastic model of the WILDHABER-NOVIKOV gears were made of Araldite CT200 in this investigation. For both the many teeth gear and the few teeth gear segments, External gears of all addendum type WILDHABER-NOVIKOV gear and the involute gear were tested. Included were the models with various profile raddi at the same pressure angle 20 .deg. and module 13.5. The flank stresses and fillet stresses of these gears were observed in each case and compared with those of gears. From this investigation, the following results were obtained. A. In the case of having many teeth gear: As the profile radius is increased, the fillet stresses of the WILDHABER-NOVIKOV gear become the same or less than that of the INVOLUTE gea, and the flank stress becomes smaller than that of the INVOLUTE gear. Therefore the better design condition is satisfied with a large profile radius. B. IN the case of having a few teeth gear: As the profile radius is increased the flank stress of WILDHABER-NOVIKOV gear becomes smaller than that of the INVOLUTE gear, but the fillet stresses become larger than that of the INVOLUTE gear. Therefore the larger design condition is satisfied with small profile radius.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.21 no.6
• /
• pp.749-754
• /
• 2011

In this study, we propose the design of optimized pRBFNNs-based face recognition system using two-dimensional Image and ASM algorithm. usually the existing 2 dimensional face recognition methods have the effects of the scale change of the image, position variation or the backgrounds of an image. In this paper, the face region information obtained from the detected face region is used for the compensation of these defects. In this paper, we use a CCD camera to obtain a picture frame directly. By using histogram equalization method, we can partially enhance the distorted image influenced by natural as well as artificial illumination. AdaBoost algorithm is used for the detection of face image between face and non-face image area. We can butt up personal profile by extracting the both face contour and shape using ASM(Active Shape Model) and then reduce dimension of image data using PCA. The proposed pRBFNNs consists of three functional modules such as the condition part, the conclusion part, and the inference part. In the condition part of fuzzy rules, input space is partitioned with Fuzzy C-Means clustering. In the conclusion part of rules, the connection weight of RBFNNs is represented as three kinds of polynomials such as constant, linear, and quadratic. The essential design parameters (including learning rate, momentum coefficient and fuzzification coefficient) of the networks are optimized by means of Differential Evolution. The proposed pRBFNNs are applied to real-time face image database and then demonstrated from viewpoint of the output performance and recognition rate.