• Journal of Korea Multimedia Society
• /
• v.19 no.8
• /
• pp.1310-1319
• /
• 2016

Face detection is the first step in a wide range of face applications. However, detecting faces in the wild is still a challenging task due to the wide range of variations in pose, scale, and occlusions. Recently, many deep learning methods have been proposed for face detection. However, further improvements are required in the wild. Another important issue to be considered in the face detection is the computational complexity. Current state-of-the-art deep learning methods require a large number of patches to deal with varying scales and the arbitrary image sizes, which result in an increased computational complexity. To reduce the complexity while achieving better detection accuracy, we propose a fully convolutional network-based face detection that can take arbitrarily-sized input and produce feature maps (heat maps) corresponding to the input image size. To deal with the various face scales, a multi-scale network architecture that utilizes the facial components when learning the feature maps is proposed. On top of it, we design multi-task learning technique to improve detection performance. Extensive experiments have been conducted on the FDDB dataset. The experimental results show that the proposed method outperforms state-of-the-art methods with the accuracy of 82.33% at 517 false alarms, while improving computational efficiency significantly.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.16 no.3
• /
• pp.877-893
• /
• 2022

With the development of deep learning, face inpainting has been significantly enhanced in the past few years. Although image inpainting framework integrated with generative adversarial network or attention mechanism enhanced the semantic understanding among facial components, the issues of reconstruction on corrupted regions are still worthy to explore, such as blurred edge structure, excessive smoothness, unreasonable semantic understanding and visual artifacts, etc. To address these issues, we propose a Learnable Structure Knowledge of Fusion Network (LSK-FNet), which learns a prior knowledge by edge generation network for image inpainting. The architecture involves two steps: Firstly, structure information obtained by edge generation network is used as the prior knowledge for face inpainting network. Secondly, both the generated prior knowledge and the incomplete image are fed into the face inpainting network together to get the fusion information. To improve the accuracy of inpainting, both of gated convolution and region normalization are applied in our proposed model. We evaluate our LSK-FNet qualitatively and quantitatively on the CelebA-HQ dataset. The experimental results demonstrate that the edge structure and details of facial images can be improved by using LSK-FNet. Our model surpasses the compared models on L1, PSNR and SSIM metrics. When the masked region is less than 20%, L1 loss reduce by more than 4.3%.

• Proceedings of the Korea Contents Association Conference
• /
• 2006.05a
• /
• pp.298-301
• /
• 2006

This study was conducted to observe the changes in trend of a beautiful face defined in different time frames. Affective space model was used in this study, for physical components of a face are far too complicated and undistinguishable from simple categorization of a face by facial characteristics. It is because using a facial model defined only by its physical characteristics renders a complicated and multi-dimensional space, whereas an emotional facial model enables a 2-dimensional space for easier observation and explanation. Scope of the study is focused on faces of the Korean female celebrities, looking at the media frequency, popularity rankings in well-known portal sites, and popularity ratings by reliable search organizations as of 2006. Celebrities in their 20s and early 30s were selected. In addition, to observe any trend in a beautiful face with the passage of time, the study looked at changes or trends in the last 5 years by comparing the faces of celebrities in year 2006 with their faces stored in the database since five years ago.

• Korean Journal of Cognitive Science
• /
• v.14 no.1
• /
• pp.10-10
• /
• 2003

This paper extracts the edge of main components of face with Gabor wavelets transformation in facial expression images. FCM(Fuzzy C-Means) clustering algorithm then extracts the representative feature points of low dimensionality from the edge extracted in neutral face. The feature-points of the neutral face is used as a template to extract the feature-points of facial expression images. To match point to Point feature points on an expression face against each feature point on a neutral face, it consists of two steps using a dynamic linking model, which are called the coarse mapping and the fine mapping. This paper presents an automatic extraction of feature-points by dynamic linking model based on Gabor wavelets and fuzzy C-means(FCM) algorithm. The result of this study was applied to extract features automatically in facial expression recognition based on dimension[1].

• Archives of Plastic Surgery
• /
• v.41 no.5
• /
• pp.588-593
• /
• 2014

Lymphedema is a chronic disorder characterized by lymph stasis in the subcutaneous tissue. Lymphatic fluid contains several components including hyaluronic acid and has many important properties. Over the past few years, significant research has been performed to identify an ideal tissue to implant as a filler. Because of its unique composition, fat harvested from the lymphedema tissue is an interesting topic for investigation and has significant potential for application as a filler, particularly in facial rejuvenation. Over a 36-month period, we treated and assessed 8 patients with lymphedematous limbs who concurrently underwent facial rejuvenation with lymphedema fat (LF). We conducted a pre- and postoperative satisfaction questionnaire survey and a histological assessment of the harvested LF fat. The overall mean general appearance score at an average of 6 months after the procedure was $7.2{\pm}0.5$, demonstrating great improvement. Patients reported significant improvement in their skin texture with a reading of $8.5{\pm}0.7$ and an improvement in their self-esteem. This study demonstrates that LF as an ideal autologous injectable filler is clinically applicable and easily available in patients with lymphedema. We recommend the further study and clinical use of this tissue as it exhibits important properties and qualities for future applications and research.

• Journal of the Korea Society of Computer and Information
• /
• v.26 no.11
• /
• pp.85-92
• /
• 2021

In this paper, a method for effectively detecting rotated face and rotation angle regardless of the rotation angle is proposed. Rotated face detection is a challenging task, due to the large variation in facial appearance. In the proposed polar coordinate transformation, the spatial information of the facial components is maintained regardless of the rotation angle, so there is no variation in facial appearance due to rotation. Accordingly, features such as HOG, which are used for frontal face detection without rotation but have rotation-sensitive characteristics, can be effectively used in detecting rotated face. Only the training data in the frontal face is needed. The HOG feature obtained from the polar coordinate transformed images is learned using SVM and rotated faces are detected. Experiments on 3600 rotated face images show a rotation angle detection rate of 97.94%. Furthermore, the positions and rotation angles of the rotated faces are accurately detected from images with a background including multiple rotated faces.

• Fashion & Textile Research Journal
• /
• v.14 no.6
• /
• pp.918-927
• /
• 2012

This study identified the aesthetic characteristics of the face description in late Joseon Dynasty beauty paintings. A total of 24 beauty paintings were selected as representative of the late Joseon Dynasty genre of painting works. The paintings were analyzed by the shape, color, and physiognomy of beauty trends from the components of women's faces expressed in the works of artists. The results of this study showed that the shape of the face components expressed a round, curved and thin line. Colors were expressed through Obang-sack (a traditional Korean color). Also the physiognomy of the late Joseon Dynasty's women was soft, wise, economical and brilliant. A round-forehead meant that economical and virtuous housekeeper, thin crescent shaped eyebrows denoted women of wisdom and excellent sensitivity. Single long thin eyelids and implied a women of longevity. A round curved nose were eager tobe a wise mother and a good wife. Small concave lips were desired eagerly by gentle and intelligent women. A curve face implied a subjective women of insight and good memory. In conclusion, the late Joseon Dynasty beauty paintings expressed a traditional Korean beauty face and a modern baby face. The data are useful for the aesthetic standards of modern through meaning of Korean traditional beauty.

• Proceedings of the IEEK Conference
• /
• 1999.06a
• /
• pp.583-586
• /
• 1999

In this paper, the wavelet transform is performed in the input 256$\times$256 color image and decomposes a image into low-pass and high-pass components. Since the high-pass band contains the components of three directions, edges are detected by combining three parts. After finding the position of face using the histogram of the edge component, a face region in low-pass band is cut off. Since RGB color image is sensitively affected by luminances, the image of low pass component is normalized, and a facial region is detected using face color informations. As the wavelet transform decomposes the detected face region into three layer, the dimension of input image is reduced. In this paper, we use the 3000 images of 10 persons, and KL transform is applied in order to classify face vectors effectively. FCM(Fuzzy C-Means) algorithm classifies face vectors with similar features into the same cluster. In this case, the number of cluster is equal to that of person, and the mean vector of each cluster is used as a codebook. We verify the system performance of the proposed algorithm by the experiments. The recognition rates of learning images and testing image is computed using correlation coefficient and Euclidean distance.

• Transactions of the Korean Society of Machine Tool Engineers
• /
• v.15 no.1
• /
• pp.39-48
• /
• 2006

In this paper, polymer composites based sandwich structures like beams and plates are optimised by using parametric study. The structures are composed of fibre reinforced composites for facial material and resin concrete and PVC foam for core materials. The stacking sequences and thickness of the composites are controlled as major parameters to find out the optimal condition for machine tool components. For the plate structure of machine tool bed composites-skined sandwich structure which has several ribs are proposed to enhance bending stiffnesses in two major directions at the same time. Dynamic robustness of a machine tool structure is investigated using modal analysis. From the results optimal configuration and materials for high precesion machine tools are proposed. And the plate was made of fiber reforced composite material and PVC foam.

• Maxillofacial Plastic and Reconstructive Surgery
• /
• v.33 no.2
• /
• pp.154-157
• /
• 2011

The primary purpose of cleft lip surgery should be the aesthetic and functional recovery of the facial components. Triangular flap repair is one of the most common techniques used in cleft lip surgery. In this case report, thirty patients with unilateral cleft lip had been treated using the Tennison-Randall method. The results were favorable and there have been no permanent complications.