• Journal of Korea Multimedia Society
• /
• v.24 no.9
• /
• pp.1242-1250
• /
• 2021

Detection of outliers deviating normal data distribution in high dimensional data is an important technique in many application areas. In this paper, a distance-based outlier detection method using landmarks in high dimensional data is proposed. Given normal training data, the k-means clustering method is applied for the training data in order to extract the centers of the clusters as landmarks which represent normal data distribution. For a test data sample, the distance to the nearest landmark gives the outlier score. In the experiments using high dimensional data such as images and documents, it was shown that the proposed method based on the landmarks of one-tenth of training data can give the comparable outlier detection performance while reducing the time complexity greatly in the testing stage.

• International Journal of Internet, Broadcasting and Communication
• /
• v.15 no.2
• /
• pp.31-38
• /
• 2023

This paper introduces a real-time face swapping system that enables video influencers to swap their faces with arbitrary generated face images of their choice. The system is implemented as a Django-based server that uses a REST request to communicate with the generative model,specifically the pretrained stable diffusion model. Once generated, the generated image is displayed on the front page so that the influencer can decide whether to use the generated face or not, by clicking on the accept button on the front page. If they choose to use it, both their face and the generated face are sent to the landmark extraction module to extract the landmarks, which are then used to swap the faces. To minimize the fluctuation of landmarks over time that can cause instability or jitter in the output, a temporal filtering step is added. Furthermore, to increase the processing speed the system works on a reduced set of the extracted landmarks.

• The korean journal of orthodontics
• /
• v.42 no.2
• /
• pp.56-63
• /
• 2012

Objective: To identify the right and left difference of the facial soft tissue landmarks three-dimensionally from the subjects of normal occlusion individuals. Materials and Methods: Cone-beam computed tomography (CT) scans were obtained in 48 normal occlusion adults (24 men, 24 women), and reconstructed into 3-dimensional (3D) models by using a 3D image soft ware. 3D position of 27 soft tissue landmarks, 9 midline and 9 pairs of bilateral landmarks, were identified in 3D coordination system, and their right and left differences were calculated and analyzed. Results: The right and left difference values derived from the study ranged from 0.6 to 4.6 mm indicating a high variability according to the landmarks. In general, the values showed a tendency to increase according to the lower and lateral positioning of the landmarks in the face. Overall differences were determined not only by transverse differences but also by sagittal and vertical differences, indicating that 3D evaluation would be essential in the facial soft tissue analysis. Conclusions: Means and standard deviations of the right and left difference of facial soft tissue landmarks derived from this study can be used as the diagnostic standard values for the evaluation of facial asymmetry.

• Maxillofacial Plastic and Reconstructive Surgery
• /
• v.42
• /
• pp.17.1-17.11
• /
• 2020

Background: To evaluate the facial asymmetry, three-dimensional computed tomography (3D-CT) has been used widely. This study proposed a method to quantify facial asymmetry based on 3D-CT. Methods: The normal standard group consisted of twenty-five male subjects who had a balanced face and normal occlusion. Five anatomical landmarks were selected as reference points and ten anatomical landmarks were selected as measurement points to evaluate facial asymmetry. The formula of facial asymmetry index was designed by using the distances between the landmarks. The index value on a specific landmark indicated zero when the landmarks were located on the three-dimensional symmetric position. As the asymmetry of landmarks increased, the value of facial asymmetry index increased. For ten anatomical landmarks, the mean value of facial asymmetry index on each landmark was obtained in the normal standard group. Facial asymmetry index was applied to the patients who had undergone orthognathic surgery. Preoperative facial asymmetry and postoperative improvement were evaluated. Results: The reference facial asymmetry index on each landmark in the normal standard group was from 1.77 to 3.38. A polygonal chart was drawn to visualize the degree of asymmetry. In three patients who had undergone orthognathic surgery, it was checked that the method of facial asymmetry index showed the preoperative facial asymmetry and the postoperative improvement well. Conclusions: The current new facial asymmetry index could efficiently quantify the degree of facial asymmetry from 3D-CT. This method could be used as an evaluation standard for facial asymmetry analysis.

• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.37 no.1
• /
• pp.203-215
• /
• 2017

Recently, interest in Pedestrian Navigation Service (PNS) is being increased due to the diffusion of smart phone and the improvement of location determination technology and it is efficient to use landmarks in route guidance for pedestrians due to the characteristics of pedestrians' movement and success rate of path finding. Accordingly, researches on extracting landmarks have been progressed. However, preceding researches have a limit that they only considered the difference between buildings and did not consider visual attention of maps in display of PNS. This study improves this problem by defining building attributes as local variable and global variable. Local variables reflect the saliency of buildings by representing the difference between buildings and global variables reflects the visual attention by representing the inherent characteristics of buildings. Also, this study considers the connectivity of network and solves the overlapping problem of landmark candidate groups by network voronoi diagram. To extract landmarks, we defined building attribute data based on preceding researches. Next, we selected a choice point for pedestrians in pedestrian network data, and determined landmark candidate groups at each choice point. Building attribute data were calculated in the extracted landmark candidate groups and finally landmarks were extracted by principal component analysis. We applied the proposed method to a part of Gwanak-gu, Seoul and this study evaluated the extracted landmarks by making a comparison with labels and landmarks used by portal sites such as the NAVER and the DAUM. In conclusion, 132 landmarks (60.3%) among 219 landmarks of the NAVER and the DAUM were extracted by the proposed method and we confirmed that 228 landmarks which there are not labels or landmarks in the NAVER and the DAUM were helpful to determine a change of direction in path finding of local level.

• The Journal of Korean Academy of Prosthodontics
• /
• v.52 no.3
• /
• pp.222-232
• /
• 2014

Purpose: This study aims to investigate if 2D analysis method is applicable to analysis of CBCT by comparing measuring points of CBCT with those of Adjusted 2D Lateral Cephalogram (Adj-Ceph) with magnification adjusted to 100% and finding out at which landmarks the difference in position appear. Materials and methods: CBCT data and Adj-Ceph (100% magnification) data from 50 adult patients have been extracted as research objects, and the horizontal (Y axis) and vertical (Z axis) coordinates of landmarks were compared. Landmarks have been categorized into 4 groups by the position and whether they are bilaterally overlapped. Paired t-test was used to compare differences between Adj-Ceph and CBCT. Results: Significant difference was found at 11 landmarks including Group B (S, Ar, Ba, PNS), Group C (Po, Or, Hinge axis, Go) and Group D (U1RP, U6CP, L6CP) in the horizontal (Y) axis while all the landmarks in vertical (Z) axis showed significant difference (P<.05). As a result of landmark difference analysis, a meaningful difference with more than 1 mm at 13 landmarks were indentifed in the horizontal axis. In the vertical axis, significant difference over 1 mm was detected from every landmark except Sella. Conclusion: Using the conventional lateral cephalometric measurements on CBCT is insufficient. A new 3D analysis or a modified 2D analysis adjusted on 19 landmarks of the vertical axis and 13 of the horizontal axis are needed when implementing CBCT diagnosis.

• Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography
• /
• v.29 no.4
• /
• pp.413-420
• /
• 2011

This study is to extract landmark buildings for pedestrian navigation from the existing spatial data sets automatically. At first, we defined candidates for landmark based on sight of pedestrian, then extracted final landmark by evaluating attributes of each candidate. The attribute is evaluated with relative or absolute criteria depending on the nature of each attribute. Landmarks extracted through the proposed method are compared to existing landmarks for vehicle and assessment of the validity and the applicability is performed. As a result, extracted Landmarks are expected to help guiding pedestrian effectively.

• Journal of Institute of Control, Robotics and Systems
• /
• v.7 no.9
• /
• pp.758-765
• /
• 2001

This paper proposes a method to automatically extract stable visual landmarks from sensory data. Given a 2D occupancy map, a mobile robot first extracts vertical line features which are distinct and on vertical planar surfaces, because they are expected to be observed reliably from various viewpoints. Since the feature information such as position and length includes uncertainty due to errors of vision and motion, the robot then reduces the uncertainty by matching the planar surface containing the features to the map. As a result, the robot obtains modeled stable visual landmarks from extracted features. This extraction process is performed on-line to adapt to an actual changes of lighting and scene depending on the robot’s view. Experimental results in various real scenes show the validity of the proposed method.

• Journal of Sensor Science and Technology
• /
• v.14 no.4
• /
• pp.278-285
• /
• 2005

This paper describes techniques of camera calibration and artificial landmarks detection for the automatic charging of a mobile robot, equipped with a fish-eye camera in the direction of its operation for movement or surveillance purposes. For its identification from the surrounding environments, three landmarks employed with infrared LEDs, were installed at the charging station. When the robot reaches a certain point, a signal is sent to the LEDs for activation, which allows the robot to easily detect the landmarks using its vision camera. To eliminate the effects of the outside light interference during the process, a difference image was generated by comparing the two images taken when the LEDs are on and off respectively. A fish-eye lens was used for the vision camera of the robot but the wide-angle lens resulted in a significant image distortion. The radial lens distortion was corrected after linear perspective projection transformation based on the pin-hole model. In the experiment, the designed system showed sensing accuracy of ${\pm}10$ mm in position and ${\pm}1^{\circ}$ in orientation at the distance of 550 mm.

• Journal of Korean Society of Industrial and Systems Engineering
• /
• v.35 no.4
• /
• pp.24-32
• /
• 2012

Of many approaches to reduce motion analysis errors, the compensation method of anatomical landmarks estimates the position of anatomical landmarks during motion. The method models the position of anatomical landmarks with joint angle or skin marker displacement using the data of the so-called dynamic calibration in which anatomical landmark positions are calibrated in ad hoc motions. Then the anatomical landmark positions are calibrated in target motions using the model. This study applies the compensation methods with joint angle and skin marker displacement to three lower extremity motions (walking, sit-to-stand/stand-to-sit, and step up/down) in ten healthy males and compares their performance. To compare the performance of the methods, two sets of kinematic variables were calculated using different two marker clusters, and the difference was obtained. Results showed that the compensation method with skin marker displacement had less differences by 30~60% compared to without compensation. And, it had significantly less difference in some kinematic variables (7 of 18) by 25~40% compared to the compensation method with joint angle. This study supports that compensation with skin marker displacement reduced the motion analysis STA errors more reliably than with joint angle in lower extremity motion analysis.