• The korean journal of orthodontics
• /
• v.51 no.6
• /
• pp.397-406
• /
• 2021

Objective: The purpose of this study was to evaluate the effects of self-ligating brackets (SBs) and other factors that influence orthodontic treatment outcomes. Methods: This two-armed cohort study included consecutively treated patients in a private practice. The patients were asked to choose between SBs and conventional brackets (CBs); if any patient did not have a preference, he or she was randomly allocated to the CB or SB group. All patients were treated using an identical archwire sequence. Evaluated parameters were as follows: treatment duration, number of bracket failures, poor oral hygiene, poor elastic wear, extraction, use of orthodontic mini-implants (OMI), OMI failure, American Board of Orthodontics (ABO) Discrepancy Index (DI), arch length discrepancy, and ABO Cast-Radiograph Evaluation (CRE) score. Stepwise regression analysis was performed to generate the equation for prediction of the CRE. Results: The final sample comprised 134 patients with an average age of 22.73 years. The average DI, CRE, and treatment duration were 21.81, 14.25, and 28.63 months, respectively. Analysis of covariance showed a significant difference in CRE between the CB and SB groups after adjusting for the effects of confounding variables. Stepwise regression analysis using four variables, namely extraction, SB use, poor elastic wear, and additional appliance use, could explain only 25.2% of the variance in the CRE. Conclusions: Although the CRE was significantly better for CBs than for SBs, the clinical significance of this result seems to be limited. Extraction, SB use, poor elastic wear, and additional appliance use may have significant effects on treatment outcomes.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.15 no.6
• /
• pp.1981-1995
• /
• 2021

When a person tries to conceal emotions, real emotions will manifest themselves in the form of micro-expressions. Research on facial micro-expression recognition is still extremely challenging in the field of pattern recognition. This is because it is difficult to implement the best feature extraction method to cope with micro-expressions with small changes and short duration. Most methods are based on hand-crafted features to extract subtle facial movements. In this study, we introduce a method that incorporates optical flow and deep learning. First, we take out the onset frame and the apex frame from each video sequence. Then, the motion features between these two frames are extracted using the optical flow method. Finally, the features are inputted into an improved MobileNetV2 model, where SVM is applied to classify expressions. In order to evaluate the effectiveness of the method, we conduct experiments on the public spontaneous micro-expression database CASME II. Under the condition of applying the leave-one-subject-out cross-validation method, the recognition accuracy rate reaches 53.01%, and the F-score reaches 0.5231. The results show that the proposed method can significantly improve the micro-expression recognition performance.

• Annual Conference of KIPS
• /
• 2021.11a
• /
• pp.704-707
• /
• 2021

Pigs' aggressive behavior represents one of the common issues that occur inside pigpens and which harm pigs' health and welfare, resulting in a financial burden to farmers. Continuously monitoring several pigs for 24 hours to identify those behaviors manually is a very difficult task for pig caretakers. In this study, we propose a lightweight video-based approach for monitoring pigs' aggressive behavior that can be implemented even in small-scale farms. The proposed system receives sequences of frames extracted from an RGB video stream containing pigs and uses MnasNet with a DM value of 0.5 to extract image features from pigs' ROI identified by predefined annotations. These extracted features are then forwarded to a lightweight LSTM to learn temporal features and perform behavior recognition. The experimental results show that our proposed model achieved 0.92 in recall and F1-score with an execution time of 118.16 ms/sequence.

• ETRI Journal
• /
• v.46 no.4
• /
• pp.671-682
• /
• 2024

Violence can be committed anywhere, even in crowded places. It is hence necessary to monitor human activities for public safety. Surveillance cameras can monitor surrounding activities but require human assistance to continuously monitor every incident. Automatic violence detection is needed for early warning and fast response. However, such automation is still challenging because of low video resolution and blind spots. This paper uses ResNet50v2 and the gated recurrent unit (GRU) algorithm to detect violence in the Movies, Hockey, and Crowd video datasets. Spatial features were extracted from each frame sequence of the video using a pretrained model from ResNet50V2, which was then classified using the optimal trained model on the GRU architecture. The experimental results were then compared with wavelet feature extraction methods and classification models, such as the convolutional neural network and long short-term memory. The results show that the proposed combination of ResNet50V2 and GRU is robust and delivers the best performance in terms of accuracy, recall, precision, and F1-score. The use of ResNet50V2 for feature extraction can improve model performance.

• Journal of Intelligence and Information Systems
• /
• v.25 no.1
• /
• pp.163-177
• /
• 2019

As smartphones are getting widely used, human activity recognition (HAR) tasks for recognizing personal activities of smartphone users with multimodal data have been actively studied recently. The research area is expanding from the recognition of the simple body movement of an individual user to the recognition of low-level behavior and high-level behavior. However, HAR tasks for recognizing interaction behavior with other people, such as whether the user is accompanying or communicating with someone else, have gotten less attention so far. And previous research for recognizing interaction behavior has usually depended on audio, Bluetooth, and Wi-Fi sensors, which are vulnerable to privacy issues and require much time to collect enough data. Whereas physical sensors including accelerometer, magnetic field and gyroscope sensors are less vulnerable to privacy issues and can collect a large amount of data within a short time. In this paper, a method for detecting accompanying status based on deep learning model by only using multimodal physical sensor data, such as an accelerometer, magnetic field and gyroscope, was proposed. The accompanying status was defined as a redefinition of a part of the user interaction behavior, including whether the user is accompanying with an acquaintance at a close distance and the user is actively communicating with the acquaintance. A framework based on convolutional neural networks (CNN) and long short-term memory (LSTM) recurrent networks for classifying accompanying and conversation was proposed. First, a data preprocessing method which consists of time synchronization of multimodal data from different physical sensors, data normalization and sequence data generation was introduced. We applied the nearest interpolation to synchronize the time of collected data from different sensors. Normalization was performed for each x, y, z axis value of the sensor data, and the sequence data was generated according to the sliding window method. Then, the sequence data became the input for CNN, where feature maps representing local dependencies of the original sequence are extracted. The CNN consisted of 3 convolutional layers and did not have a pooling layer to maintain the temporal information of the sequence data. Next, LSTM recurrent networks received the feature maps, learned long-term dependencies from them and extracted features. The LSTM recurrent networks consisted of two layers, each with 128 cells. Finally, the extracted features were used for classification by softmax classifier. The loss function of the model was cross entropy function and the weights of the model were randomly initialized on a normal distribution with an average of 0 and a standard deviation of 0.1. The model was trained using adaptive moment estimation (ADAM) optimization algorithm and the mini batch size was set to 128. We applied dropout to input values of the LSTM recurrent networks to prevent overfitting. The initial learning rate was set to 0.001, and it decreased exponentially by 0.99 at the end of each epoch training. An Android smartphone application was developed and released to collect data. We collected smartphone data for a total of 18 subjects. Using the data, the model classified accompanying and conversation by 98.74% and 98.83% accuracy each. Both the F1 score and accuracy of the model were higher than the F1 score and accuracy of the majority vote classifier, support vector machine, and deep recurrent neural network. In the future research, we will focus on more rigorous multimodal sensor data synchronization methods that minimize the time stamp differences. In addition, we will further study transfer learning method that enables transfer of trained models tailored to the training data to the evaluation data that follows a different distribution. It is expected that a model capable of exhibiting robust recognition performance against changes in data that is not considered in the model learning stage will be obtained.

• Asian-Australasian Journal of Animal Sciences
• /
• v.22 no.3
• /
• pp.313-318
• /
• 2009

TIAF1 is a TGF-${\beta}$1-induced anti-apoptotic factor that plays a critical role in blocking TNF (tumor necrosis factor) cytotoxicity in mouse fibroblasts and participates in TGF-${\beta}$-mediated growth regulation. In this study, we obtained the full-length cDNA sequence of the porcine TIAF1 gene. Real-time PCR further revealed that the TIAF1 gene was expressed at the highest level in liver and kidney with prominent expressions detected in uterus, and lower levels detected in heart, spleen, lung, stomach, small intestine, skeletal muscle and fat of Large White pigs. Sequence analysis indicated that a 6 base-pair deletion mutation existed in the exon of the TIAF1 gene between Meishan and Large White pigs. This mutation induced deletion of Gln and Val amino acids. PCR-RFLP was used to detect the polymorphism in 394 pigs of a "Large White${\times}$Meishan" $F_{2}$ resource population and four purebred pig populations. The frequencies of the A allele (with a 6 bp deletion) were dominant in Chinese Meishan and Bamei pigs, and the frequencies of the B allele (no 6 bp deletion) were dominant in Large White and Landrace pigs. Association analyses revealed that the deletion mutation had highly significant associations (p<0.01) with meat marbling score of the thorax-waist longissimus dorsi (LD) muscle (MM1) and intramuscular fat percentage (IMF), and significant associations (p<0.05) with carcass length (CL). The results presented here supply evidence that the 6 bp deletion mutation in the TIAF1 gene affects porcine meat quality and provides useful information for further porcine breeding.

• Investigative Magnetic Resonance Imaging
• /
• v.23 no.3
• /
• pp.210-219
• /
• 2019

Purpose: The purpose of this study was to investigate if double inversion recovery (DIR) imaging can have a role in the evaluation of brain ischemia, compared with diffusion-weighted imaging (DWI) and fluid-attenuated inversion recovery (FLAIR) imaging. Materials and Methods: Sixty-seven patients within 48 hours of onset, underwent MRI scans with FLAIR, DWI with b-value of 0 (B0) and $1000s/mm^2$, and DIR sequences. Patients were categorized into four groups: within three hours, three to six hours, six to 24 hours, and 24 to 48 hours after onset. Lesion-to-normal ratio (LNR) value was calculated and compared among all sequences within each group, by the Friedman test and conducted among all groups, for each sequence by the Kruskal-Wallis test. In qualitative assessment, signal intensity changes of DIR, B0, and FLAIR based on similarity with DWI and image quality of each sequence, were graded on a 3-point scale, respectively. Scores for detectability of lesions were compared by the McNemar's test. Results: LNR values from DWI were higher than DIR, but not statistically significant in all groups (P > 0.05). LNR values of DIR were significantly higher than FLAIR within 24 hours of onset (P < 0.05). LNR values were significantly different between, before, and after six hours onset time for DIR (P = 0.016), B0 (P = 0.008), and FLAIR (P = 0.018) but not for DWI (P = 0.051). Qualitative analysis demonstrated that detectability of DIR was higher, compared to that of FLAIR within 4.5 hours and six hours of onset (P < 0.05). Also, the DWI quality score was lower than that of DIR, particularly relative to infratentorial lesions. Conclusion: DIR provides higher detectability of hyperacute brain ischemia than B0 and FLAIR, and does not suffer from susceptibility artifact, unlike DWI. So, DIR can be used to replace evaluation of the FLAIR-DWI mismatch.

• Journal of the Korean Institute of Landscape Architecture
• /
• v.15 no.1
• /
• pp.39-67
• /
• 1987

The purpose of this thesis is to suggest objective basic data for the environmental design through the quantitative analysis of the visual quality included in the physical environment of forest landscape. For this, landscape values of forest landscape have been evaluated by using the Iverson method, the images structure of forest landscape's main utilizing space have been analysed by the factor analysis algorithm, degree of visual preferences have been pleasured mainly by questionnaries and SBE method, and finally these thesis can be summarized as fallow LCP with high values of Iverson factors I and IV yield high landscape value. Specifically, Iverson factor IV has been found to play the dominant. For all experimental points, significant seasonal variations in S.D. scale values have been observed. In natural parks, where artificial structures are complementary to the natural landscape, main factors of image are S.D. scales such as the visual sequence, the formal simplicity of structures, the emphasis, the unification of heterogeneous factors and the assimilation. Factors covering the spatial image of natural parks have been found to be the overall evaluation, the individual characteristics, the tidiness, the potentiality, the dignity, the intimacy and the space volume. For all seasons, factors such as the individual characteristics, the dignity, the tidiness, the potentiality, yield high factor scores. As for factors determining the degree of visual preference, variables such as the summit, the skyline, rocks, the water and the degree of natural destruction by artificial structures yield high values for all seasons.

• Journal of the Korea Convergence Society
• /
• v.8 no.5
• /
• pp.79-85
• /
• 2017

The purpose of this research was to evaluate the clinical and the instrument of convergence utility of transient elastography (FibroScan(R):electromagnetic wave) in diagnosing and treating liver ailments through a comparison and an analysis between liver function blood test and transient elastography (FibroScan(R)) in patients with chronic hepatitis B virus infection. Of all the patients with chronic hepatitis B virus infection who visited clinic B in Daejeon City between July 1, 2015, and February 28, 2016, 75 who underwent a FibroScan(R) test were selected for this study. Their laboratory and liver function test results were compared for a correlation analysis before constructing an ROC (Receiver Operation Characteristic) curve. Aspartate aminotransferase (AST) and alanine aminotransferase (ALT) levels were 0.572 and 0.502, respectively, and showed highest correlation with fibrosis score, with statistical significance (p<0.000). Gamma glutamyltranspeptidase, total bilirubin, and alkaline phosphatase levels also showed relatively significant correlations in this order of sequence, while -fetoprotein and total protein levels did not show any statistically significant values. Albumin (-0.449) and platelet levels (-0.373) showed negative correlations with each other and no correlation with fibrosis score (p < 0.000). As liver fibrosis worsened, the accuracy of the ROC curve increased. At the F4 stage, which is the cirrhotic stage, the largest area under the curve was observed. FibroScan(R) showed significant correlation with the ALT (serum glutamic pyruvic transaminase) and AST (serum glutamic oxaloacetic transaminase) levels in the liver function test, which is a routine test for patients with chronic liver ailments. This implies that fibrosis correlates with liver inflammation severity.

• Science of Emotion and Sensibility
• /
• v.13 no.1
• /
• pp.251-258
• /
• 2010

'Naturalness' is the important "ness" which is a key factor in image quality assessment. 'Naturalness' is a representive factor depending on the context of the image which arouses different emotions. The Image Quality Circle was split into two steps. The first step is predicting the visual perceptual attribute which are lightness, colourfulness, hue and contrast. The next step is SSE which is dependent to image contents. In this study the image contents are grouped in genres. The images were rendered using four different colour attributes which are lightness, contrast, colourfulness and hue. Using a scale, the score of image quality and SSE was asked to each participant for all rendered images. A seven-point category scale of increasing amount of "ness" is used as a quantitative adjectives sequence. The image quality model was built by combining the SSEs for each scene. The SSEs, where vividness is common, are considered as independent variables to predict the image quality score. Then the vividness model was built using colour attributes as variables to predict the vividness of each scene (genre). Vividness is an important factor of naturalness which the meaning is different for all scenes that links the naturalness and image quality. The vividness meaning was different for each scene (genre). Therefore, the colour attributes that express the vividness would depend on the image content.