• Korean Journal of Remote Sensing
• /
• v.40 no.1
• /
• pp.115-122
• /
• 2024

Multi-object tracking (MOT) is a vital component in understanding the surrounding environments. Previous research has demonstrated that MOT can successfully detect and track surrounding objects. Nonetheless, inaccurate classification of the tracking objects remains a challenge that needs to be solved. When an object approaching from a distance is recognized, not only detection and tracking but also classification to determine the level of risk must be performed. However, considering the erroneous classification results obtained from the detection as the track class can lead to performance degradation problems. In this paper, we discuss the limitations of classification in tracking under the classification uncertainty of the detector. To address this problem, a class update module is proposed, which leverages the class uncertainty estimation of the detector to mitigate the classification error of the tracker. We evaluated our approach on the VisDrone-MOT2021 dataset,which includes multi-class and uncertain far-distance object tracking. We show that our method has low certainty at a distant object, and quickly classifies the class as the object approaches and the level of certainty increases.In this manner, our method outperforms previous approaches across different detectors. In particular, the You Only Look Once (YOLO)v8 detector shows a notable enhancement of 4.33 multi-object tracking accuracy (MOTA) in comparison to the previous state-of-the-art method. This intuitive insight improves MOT to track approaching objects from a distance and quickly classify them.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.13 no.11
• /
• pp.5196-5205
• /
• 2012

Purpose: The Purpose of this study were to compare the level of perception and to identify factors associated with perception on patient safety culture by nurses and hygienists. Method: The data were collected from september to December, 2010 using Hospital survey on patient safety culture questionnaires. The subjects were 399 Nurses, hygienists, recruited from the hospital in Busan & Kyungnam. The collected data were analyzed using SPSS descriptive statistics, mean and standard deviation, t-test and ANOVA, Spearman rank coefficient. Result: The perception level of nurses on patient safety culture was 3.48. In case of hygienists, the level was 3.51. Compared to nurses, hygienists showed a significantly difference on the items "Staff arrangement"(t=2.841, p<.01) and "Administator attitude"(t=-2.471, p<.05), "Feedback and communication in accident"(t=-3.356, p<.01). Nurses and hygienists' age and career, working hour per week were identified as factor associated with patient safety culture. Conclusion: The perception level of hospital health providers on patient safety culture was moderate. and identified factors associated with patient safety culture were age and career, working hour per week.

• Journal of Korea Water Resources Association
• /
• v.52 no.12
• /
• pp.975-984
• /
• 2019

As an essential prerequisite for systematic and integrated management of river water, it is necessary to secure the basic data such as discharge supplied to the river and released from the river. Under the current permit system for river water use, 59.1% of licensed facilities were found to have no discharge meters in 2017, especially for agricultural water, which makes it difficult to secure reliable data as a large portion of the reports are voluntarily reported by users. In this study, the indirect discharge measurement method of calculating the discharge through the power usage of the pumping station was applied to secure reliable discharge data. In particular, focusing on the fact that the discharge calculated by the power usage method differed with the actual discharge according to the level of the river, the study was conducted on improving the power usage method reflecting the river water level and improving the accuracy of discharge data. Analysis of the discharge calculated using the power usage method considering river water level using the correlation analysis method such as regression analysis, percent difference, root mean square error etc. confirmed that the results are not high compared to the conventional power usage method, but are slightly more approximated to the actual discharge. Therefore, although reliable discharge data can be obtained from the existing power usage method, it is expected that more accurate data on intaking water of river water can be obtained if the improved power usage method is used at points where the variation in the water level of the river is large.

• Journal of Korean Neurosurgical Society
• /
• v.57 no.2
• /
• pp.88-93
• /
• 2015

Objective : Elevated cellular retinoic acid binding protein-I (CRABP-I) is thought to be related to the abnormal proliferation and migration of smooth muscle cells (SMCs). Accordingly, a higher CRABP-I level could cause disorganized vessel walls by causing immature SMC phenotypes and altering extracellular matrix proteins which could result in vulnerable arterial walls with inadequate responses to hemodynamic stress. We hypothesized that elevated CRABP-I level in the cerebrospinal fluid (CSF) could be related to subarachnoid hemorrhage (SAH). Moreover, we also extended this hypothesis in patients with vascular malformation according to the presence of hemorrhage. Methods : We investigated the CSF of 26 patients : SAH, n=7; unruptured intracranial aneurysm (UIA), n=7; arteriovenous malformation (AVM), n=4; cavernous malformation (CM), n=3; control group, n=5. The optical density of CRABP-I was confirmed by Western blotting and presented as mean${\pm}$standard error of the measurement. Results : CRABP-I in SAH ($0.33{\pm}0.09$) was significantly higher than that in the UIA ($0.12{\pm}0.01$, p=0.033) or control group ($0.10{\pm}0.01$, p=0.012). Hemorrhage presenting AVM (mean 0.45, ranged 0.30-0.59) had a higher CRABP-I level than that in AVM without hemorrhage presentation (mean 0.16, ranged 0.14-0.17). The CRABP-I intensity in CM with hemorrhage was 0.21 and 0.31, and for CM without hemorrhage 0.14. Overall, the hemorrhage presenting group (n=11, $0.34{\pm}0.06$) showed a significantly higher CRABP-I intensity than that of the non-hemorrhage presenting group (n= 10, $0.13{\pm}0.01$, p=0.001). Conclusion : The results suggest that elevated CRABP-I in the CSF could be related with aneurysm rupture. Additionally, a higher CRABP-I level seems to be associated with hemorrhage development in vascular malformation.

• Journal of Korean Society of Transportation
• /
• v.33 no.5
• /
• pp.497-507
• /
• 2015

Nonrecurrent congestion of freeway was primarily caused by incident. The main cause of incident was known as a traffic accident. Therefore, accurate prediction of traffic incident clearance time is very important in accident management. Traffic accident data on freeway during year 2008 to year 2014 period were analyzed for this study. KNN(K-Nearest Neighbor) algorithm was hired for developing incident clearance time prediction model with the historical traffic accident data. Analysis result of accident data explains the level of accident significantly affect on the incident clearance time. For this reason, incident clearance time was categorized by accident level. Data were sorted by classification of traffic volume, number of lanes and time periods to consider traffic conditions and roadway geometry. Factors affecting incident clearance time were analyzed from the extracted data for identifying similar types of accident. Lastly, weight of detail factors was calculated in order to measure distance metric. Weight was calculated with applying standard method of normal distribution, then incident clearance time was predicted. Prediction result of model showed a lower prediction error(MAPE) than models of previous studies. The improve model developed in this study is expected to contribute to the efficient highway operation management when incident occurs.

• Journal of Korean Academy of Nursing
• /
• v.30 no.3
• /
• pp.549-559
• /
• 2000

The purpose of this study was to investigate the correlation between the knowledge and educational needs related to recurrent in coronary artery bypass graft patients as a basis to provide an individual nursing education for the population. The subjects consisted of 110 patients who had coronary artery bypass graft(CABG) at Asan Medical Center in Seoul and Sechong hospital in Buchon. Data was obtained from a knowledge questionnaire and a learning needs questionnaire between November 1998 and February 1999. Data were analyzed using SAS program for Wilcoxon rank sum test and Spearman correlation coefficient. The results were as follows : 1. With regard to the 18 items to measure knowledge, the mean (median) of items 'don't know' was 4.9(4) items. The mean (median) of items answered wrong was 3.2(3) items. The number of items answered 'don't know' tend to show higher in those who had less education, blue color jobs and myocardiac infarction history than in their counter parts. There were higher frequency of items answered 'don't know' in those who had no hypertension 2. With regard to the level of knowledge by questionnaire about CABG, The most "I dont know" (59.1%) highly response was 'He has to be treated with anticoagulant drug to prevent revasculized vessel from obstructing.' The seond highest response (56.4%) was 'If you were hypotensive, the coronary attack would collapse. 'During the hospitalized day, the patient has complete bedrest.' The highest error probability was cholesterol has not to intake.', 'After surgery, the sexual life is need controlled for 1 year. 3. The mean of educational needs was 3.38. With regard to the level of learning needs by sentence about CABG, 'Food that benefit heart disease', 'Recurrence possibility of heart disease', 'Management method of operation site', 'Risk symptom that visit hospital or report immediately' were higher than other sentenses. With regard to the level of learning needs by factor 'food(5 items)', 'disease(9 items)' and 'exercise(3 items)' showed the highest than other factors. The educational needs by patients characteristics tend to show higher in males, under the age of 49, middle or high school degree, previous experience of admission with coronary artery disease, history of myocardial infarction, expierience of PTCA, history of cerebro-vascular accident, previous expierience of smoking than in their counter parts. 4. The number of items answered 'don't know', wrong and correct weren't correlated with the level educational needs. As the results, the number of items answered 'don't know' tend to show higher in those who had less educated, blue color jobs and myocardiac infarction history than in their counter parts. There were higher frequency of items answered 'don't know' in those who had no hypertension .There were higher frequency of items answered 'don't know' on anti-thrombolitic theraphy, hypotension and pain relief. Also there were higher frequency of items answered wrong on bed rest period, cholesterol intake, and sexual life. Educational needs were higher in young age group, had previous experience of procesure and history of other disease. And when we educate CABG patients, education for diet, recurrence possibility of disease, management methods of operation site and risk symptom should be emphasized.

• The Journal of the Acoustical Society of Korea
• /
• v.39 no.5
• /
• pp.497-504
• /
• 2020

Constructing speaker embeddings in speaker verification is an important issue. In general, a self-attention mechanism has been applied for speaker embedding encoding. Previous studies focused on training the self-attention in a high-level layer, such as the last pooling layer. In this case, the effect of low-level layers is not well represented in the speaker embedding encoding. In this study, we propose Masked Cross Self-Attentive Encoding (MCSAE) using ResNet. It focuses on training the features of both high-level and low-level layers. Based on multi-layer aggregation, the output features of each residual layer are used for the MCSAE. In the MCSAE, the interdependence of each input features is trained by cross self-attention module. A random masking regularization module is also applied to prevent overfitting problem. The MCSAE enhances the weight of frames representing the speaker information. Then, the output features are concatenated and encoded in the speaker embedding. Therefore, a more informative speaker embedding is encoded by using the MCSAE. The experimental results showed an equal error rate of 2.63 % using the VoxCeleb1 evaluation dataset. It improved performance compared with the previous self-attentive encoding and state-of-the-art methods.

• Journal of Korea Water Resources Association
• /
• v.54 no.12
• /
• pp.1285-1294
• /
• 2021

In recent years, as rainfall is concentrated and rainfall intensity increases worldwide due to climate change, the scale of flood damage is increasing. Rainfall of a previously unobserved magnitude falls, and the rainy season lasts for a long time on record. In particular, these damages are concentrated in ASEAN countries, and at least 20 million people among ASEAN countries are affected by frequent flooding due to recent sea level rise, typhoons and torrential rain. Korea supports the domestic flood warning system to ASEAN countries through various ODA projects, but the communication network is unstable, so there is a limit to the central control method alone. Therefore, in this study, an artificial intelligence-based flood prediction model was developed to develop an observation station that can observe water level and rainfall, and even predict and warn floods at once at one observation station. Training, validation and testing were carried out for 0.5, 1, 2, 3, and 6 hours of lead time using the rainfall and water level observation data in 10-minute units from 2009 to 2020 at Junjukbi-bridge station of Seolma stream. LSTM was applied to artificial intelligence algorithm. As a result of the study, it showed excellent results in model fit and error for all lead time. In the case of a short arrival time due to a small watershed and a large watershed slope such as Seolma stream, a lead time of 1 hour will show very good prediction results. In addition, it is expected that a longer lead time is possible depending on the size and slope of the watershed.

• Journal of IKEEE
• /
• v.25 no.4
• /
• pp.782-785
• /
• 2021

In this paper, we propose a personalized cooling management system with thermal imaging camera. The proposed equipment uses a thermal imaging camera to control the amount of cold air and the system according to the difference between the user's skin temperature before and after the procedure. When the skin temperature is abnormally low, the cold air supply is cut off to prevent the possibility of a safety accident. It is economical by replacing the skin temperature sensor with a thermal imaging camera temperature measurement, and it can be visualized because the temperature can be checked with the thermal image. In addition, the proposed equipment improves the sensitivity of the sensor that measures the distance to the skin by calculating the focal length by using a dual laser pointer for the safety of a personalized cooling management system to which a thermal imaging camera is applied. In order to evaluate the performance of the proposed equipment, it was tested in an externally accredited testing institute. The first measured temperature range was -100℃~-160℃, indicating a wider temperature range than -150~-160℃(cryo generation/USA), which is the highest level currently used in the field. In addition, the error was measured to be ±3.2%~±3.5%, which showed better results than ±5%(CRYOTOP/China), which is the highest level currently used in the field. The second measured distance accuracy was measured as below ±4.0%, which was superior to ±5%(CRYOTOP/China), which is the highest level currently used in the field. Third, the nitrogen consumption was confirmed to be less than 0.15 L/min at the maximum, which was superior to the highest level of 6 L/min(POLAR BEAR/USA) currently used in the field. Therefore, it was determined that the performance of the personalized cooling management system applied with the thermal imaging camera proposed in this paper was excellent.

• Journal of Gastric Cancer
• /
• v.22 no.2
• /
• pp.120-134
• /
• 2022

Purpose: This study aimed to identify prognostic factors for patients with distant lymph node-involved gastric cancer (GC) using a machine learning algorithm, a method that offers considerable advantages and new prospects for high-dimensional biomedical data exploration. Materials and Methods: This study employed 79 features of clinical pathology, laboratory tests, and therapeutic details from 289 GC patients whose distant lymphadenopathy was presented as the first episode of recurrence or metastasis. Outcomes were measured as any-cause death events and survival months after distant lymph node metastasis. A prediction model was built based on possible outcome predictors using a random survival forest algorithm and confirmed by 5×5 nested cross-validation. The effects of single variables were interpreted using partial dependence plots. A contour plot was used to visually represent survival prediction based on 2 predictive features. Results: The median survival time of patients with GC with distant nodal metastasis was 9.2 months. The optimal model incorporated the prealbumin level and the prothrombin time (PT), and yielded a prediction error of 0.353. The inclusion of other variables resulted in poorer model performance. Patients with higher serum prealbumin levels or shorter PTs had a significantly better prognosis. The predicted one-year survival rate was stratified and illustrated as a contour plot based on the combined effect the prealbumin level and the PT. Conclusions: Machine learning is useful for identifying the important determinants of cancer survival using high-dimensional datasets. The prealbumin level and the PT on distant lymph node metastasis are the 2 most crucial factors in predicting the subsequent survival time of advanced GC.