• The Journal of the Acoustical Society of Korea
• /
• v.18 no.6
• /
• pp.49-56
• /
• 1999

In many continuous speech recognition systems based on HMMs, decision tree-based state tying has been used for not only improving the robustness and accuracy of context dependent acoustic modeling but also synthesizing unseen models. To construct the phonetic decision tree, standard method performs one-level pruning using just single Gaussian triphone models. In this paper, two novel approaches, two-level decision tree and multi-mixture decision tree, are proposed to get better performance through more accurate acoustic modeling. Two-level decision tree performs two level pruning for the state tying and the mixture weight tying. Using the second level, the tied states can have different mixture weights based on the similarities in their phonetic contexts. In the second approach, phonetic decision tree continues to be updated with training sequence, mixture splitting and re-estimation. Multi-mixture Gaussian as well as single Gaussian models are used to construct the multi-mixture decision tree. Continuous speech recognition experiment using these approaches on BN-96 and WSJ5k data showed a reduction in word error rate comparing to the standard decision tree based system given similar number of tied states.

• Journal of the Institute of Electronics Engineers of Korea CI
• /
• v.43 no.2 s.308
• /
• pp.21-32
• /
• 2006

The license plate recognition (LPR) unit consists of the following core components: plate region segmentation, individual character extraction, and character recognition. Out of the above three components, accuracy in the performance of plate region segmentation determines the overall recognition rate of the LPR unit. This paper proposes an algorithm for segmenting the license plate region on the front or rear of a vehicle in a fast and accurate manner. In the case of the proposed algorithm images are captured on the spot where unmanned monitoring of illegal parking and stowage is performed with a variety of roadway environments taken into account. As a means of enhancing the segmentation performance of the on-the-spot-captured images of license plate regions, the proposed algorithm uses a mathematical model for license plate colors to convert color images into digital data. In addition, this algorithm uses Gaussian smoothing and double threshold to eliminate image noises, one-pass boundary tracing to do region labeling, and MBR to determine license plate region candidates and extract individual characters from the determined license plate region candidates, thereby segmenting the license plate region on the front or rear of a vehicle through a verification process. This study contributed to addressing the inability of conventional techniques to segment the license plate region on the front or rear of a vehicle where the frame of the license plate is damaged, through processing images in a real-time manner, thereby allowing for the practical application of the proposed algorithm.

• Journal of the korean academy of Pediatric Dentistry
• /
• v.41 no.3
• /
• pp.199-206
• /
• 2014

The number of reported traumatic injuries has been increasing in recent years, particularly that of dental trauma in children and adolescents. While the risks associated with dental trauma in this population have been increasingly realized, domestic reports on this issue seem to be insufficient. The purpose of this study was to investigate the prevalence of dental trauma and to evaluate the relevant risk factors and the level of self-recognition. 1,371 adolescents attending middle schools at Yangsan were surveyed via clinical examination and questionnaire; the results were as follows: According to the clinical examination, the overall prevalence of dental trauma was 16.8% with preponderance of males (19.2%) compared to females (13.7%) (p < 0.05). The average number of injured teeth per adolescent with a history of trauma was 1.34. Comparing the prevalence, maxillary central incisors and enamel fracture occupied the highest ranks by tooth type and mode of trauma, respectively. Class II division 1 malocclusion and overjet exceeding 8.0 mm were identified as significant risk factors (p < 0.05). The degree of self-recognition of dental trauma showed a low coincidence rate compared with results of the clinical examination. Males expressed a higher level of satisfaction toward the results of treatment for dental trauma than females (p < 0.05).

• Korean Journal of Biological Psychiatry
• /
• v.17 no.2
• /
• pp.79-85
• /
• 2010

Objectives : Trauma patients have attentional bias which enforces traumatic memories and causes cognitive errors. Understanding of such selective attention may explain many aspects of the posttraumatic stress disorder(PTSD) symptoms. Methods : We used the rapid serial visual presentation(RSVP) method to verify attentional blink in burn patients with PTSD. International affective picture system(IAPS) was used as stimuli and distracters. In the 'neutral test', patients have been presented series of pictures with human face picture as target stimuli. Each picture had 100ms interval. However the distance between target facial pictures was randomized and recognition of second facial picture accuracy was measured. In the 'stress test', the first target was stress picture which arouses patient emotions instead of the facial picture. Neutral and Stress tests were done with seven PTSD patients and 20 controls. In '85ms test' the interval was reduced to 85ms. The accuracy of recognition of second target facial picture was rated in all three tests. Eighty-five ms study was done with eighteen PTSD patients. Results : Attentional blinks were observed in 100-400ms of RSVP. PTSD patients showed increased recognition rate in the 'stress test' compared with the 'neutral test'. When presentation interval was decreased to 85 ms, PTSD patient showed decrease of attentional blink effect when target facial picture interval was 170ms. Conclusion : We found attentional blink effect could be affected by stress stimulus in burn patients. And attentional blink may be affected by stimulus interval and the character of stimulus. There may be some other specific mechanism related with selective attention in attentional blink especially with facial picture processing.

• KIPS Transactions on Software and Data Engineering
• /
• v.1 no.2
• /
• pp.127-136
• /
• 2012

Many research of hand gesture recognition based on vision system have been conducted which enable user operate various electronic devices more easily. 3D position calculation and meaningful gesture classification from similar gestures should be executed to recognize hand gesture accurately. A simple and cost effective method of 3D position calculation and gesture spotting (a task to recognize meaningful gesture from other similar meaningless gestures) is described in this paper. 3D position is achieved by calculation of two cameras relative position through pan/tilt module and a marker regardless with the placed position. Fuzzy garbage model is proposed to provide a variable reference value to decide whether the user gesture is the command gesture or not. The reference is achieved from fuzzy command gesture model and fuzzy garbage model which returns the score that shows the degree of belonging to command gesture and garbage gesture respectively. Two-stage user adaptation is proposed that off-line (batch) adaptation for inter-personal difference and on-line (incremental) adaptation for intra-difference to enhance the performance. Experiment is conducted for 5 different users. The recognition rate of command (discriminate command gesture) is more than 95% when only one command like meaningless gesture exists and more than 85% when the command is mixed with many other similar gestures.

• The Journal of The Korea Institute of Intelligent Transport Systems
• /
• v.18 no.3
• /
• pp.95-105
• /
• 2019

In recent years, image processing techniques for detecting road surface damaged spot have been actively researched. Especially, it is mainly used to acquire images through a smart phone or a black box that can be mounted in a vehicle and recognize the road surface damaged region in the image using several algorithms. In addition, in conjunction with the GPS module, the exact damaged location can be obtained. The most important technology is image processing algorithm. Recently, algorithms based on artificial intelligence have been attracting attention as research topics. In this paper, we will also discuss artificial intelligence image processing algorithms. Among them, an object detection method based on an region-based convolution neural networks method is used. To improve the recognition performance of road surface damage objects, 600 road surface damaged images and 1500 general road driving images are added to the learning database. Also, supervised learning using background object recognition method is performed to reduce false alarm and missing rate in road surface damage detection. As a result, we introduce a new method that improves the recognition performance of the algorithm to 8.66% based on average value of mAP through the same test database.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.20 no.2
• /
• pp.750-762
• /
• 2019

The importance of the call center as a contact point for the enterprise is growing. However, call centers have difficulty with their operating agents due to the agents' lack of knowledge and owing to frequent agent turnover due to downturns in the business, which causes deterioration in the quality of customer service. Therefore, through an N-bank call center case study, we developed a system to reduce the burden of keeping up business knowledge and to improve customer service quality. It is a "real-time agent advisor" system that provides agents with answers to customer questions in real time by combining AI technology for speech recognition, natural language processing, and questions & answers for existing call center information systems, such as a private branch exchange (PBX) and computer telephony integration (CTI). As a result of the case study, we confirmed that the speech recognition system for real-time call analysis and the corpus construction method improves the natural speech processing performance of the query response system. Especially with name entity recognition (NER), the accuracy of the corpus learning improved by 31%. Also, after applying the agent advisor system, the positive feedback rate of agents about the answers from the agent advisor was 93.1%, which proved the system is helpful to the agents.

• Journal of Industrial Convergence
• /
• v.20 no.12
• /
• pp.213-217
• /
• 2022

The purpose of this study was to investigate the relationship between age, physical activity (walking, strength exercise), subjective body shape recognition, and weight change with subjective health status in the population aged 30 years or older among the data of the National Health and Nutrition Examination Survey (2018) of Koreans. The subject and method of this study was a secondary analysis study using the 7th 3rd year National Health and Nutrition Examination Survey (2018). Age, walking days per week, strength training days per week, subjective health status, subjective body shape recognition, and weight change during the year were used the SPSS (p<.01). The response rate of walking for 7 days (every day) was 21.9% for those aged 30-45, 27.2% for those aged 50-64, and 26.6% for those aged 65 or older. aged 8.4% and over 65 years old 7.9% (p<.001). The subjective health status was positively correlated with age (.091), number of walking days per week (.739), number of days of strength exercise per week (.530), subjective perception of body shape (.256), and weight change over one year (.303). There was a relationship (p<.01). In this study, the age group that answered 'bad' in subjective health status was the highest at 21.4% of those aged 65 or older, followed by 15.7% of those aged 50-64 and 11.0% of those aged 30-45. This could be expected to improve subjective health status by walking and strength training. Based on these results, it was intended to provide basic data for preparing health-related programs to improve subjective health status of individuals according to age.

• Journal of the Korean Society for information Management
• /
• v.40 no.2
• /
• pp.81-114
• /
• 2023

In this study, a survey was conducted targeting specialized librarians, and the impact on the work area according to changes in the internal and external environment and policy support measures was analyzed. In this study, we tried to derive factors that affect library development and policy suggestions accordingly. As a result of the study, first, it was confirmed that 58.3% of the negative opinions in terms of the importance of library development plans were positive in recognition of the role of library status within individual institutions. Second, in order to increase the status of specialized libraries, it was found that awareness of academic research activities was necessary by recognizing the importance of major functions and roles. Third, among the comprehensive library development plans, the recognition of specialized libraries and operational evaluation was the highest in recognition of the expansion of national public information services to the public. In addition, it was confirmed that among the five-year development strategies, the policy that should be implemented first is the preference for updating the status of specialized libraries and establishing a system for investigation. Fourth, as a result of analyzing effective alternatives and improvement indicators to increase the participation rate in library operation evaluation, the weighting of the "institutional library operation evaluation" item in the evaluation item of public enterprises was the highest at 4.01 on average. Therefore, for the development of specialized libraries, it was recognized as the most urgent task to establish a system that can comprehensively grasp the current status of specialized libraries as well as active academic research and support them.

• Journal of the Korean Society of School Health
• /
• v.4 no.2
• /
• pp.193-215
• /
• 1991

This study is aimed at providing basic information applicable to setting up the education programs and strategies to prevent smoking among high school students by analysing smoking behavior and influencing factors. The samples consisted of 814 students, 557 parents and 362 teachers selected randomly from 8 high schools, one from each school district in Seoul. Date Analyses were made through Chi-Square test, Factor Analysis, One Way ANOVA, Multiple Regression, Correlation. SPSS/$PC^+$program was utilized. Smoking behavior (continuous smoking, re-smoking, ex-smoking, never smoking, daily smoking, occasional smoking) were used as dependent variables. Influencing factors (male students, habits, attitudes and knowledge toward smoking, home life, school life, juvenile delinquency, friendship, demographic parent's and teacher's recognition toward male students smoking) were used as in dependent variables. The major findings of the study are as follows : 1. The total smoking rate occupies 41.1% whereas the continuous smoking rate stands at 19.2%, re-smoking rate 9.5%, ex-smoking rate 12.4% and never smoking rate 58.9%. 2. The total smoking rate among high school students is significantly correlated with their monthly expenditures and type of school (p<0.001). The continuous smoking rate also shows the same tendency. As the length of butt get shorter, the current smoking rate increase. The duration of smoking is in proportion to its continuity. The major motive of smoking is curiousity whereas that re-smoking is to follow friend's behavior. The study shows that peer pressure is the most powerful factor influencing smoking behavior of students. Friends and fellow students encourage to pick up smoking and resume smoking even alter one stops smoking. 3. The correlationship between favorable attitudes toward smoking and the current smoking rate and its continuity is statistically significant(p<0.05, p<0.01, p<0.001). 4. The stability and harmony of family life and the current smoking rate show negative correlationship. The daily smoking amount of father is in proportion to the continuity of students' smoking. When parents are in favor of smoking, it is more likely that the experimental smoking rate increase the smoking rate increases, and vice versa. The more acceptable attitudes toward smoking among siblings is also one of the factors to increase the smoking rate and continuity (p<0.001). The more lenient the attitudes of parents toward their children's association with smoking friends, the higher the smoking rate. When students have difficulties in adjusting to school life, it is more likely that the current smoking rate and continuity increase. 5. The continuity of smoking and friendship are significantly correlated (p<0.05, p<0.01, p<0.001). 6. The continuity of smoking and juvenile delinquency are significantly correlated (p<0.001). 7. The difference in attitudes and smoking reasons of parents and students is significantly correlated to different smoking behavior (p<0.01, p<0.001). While smoking knowledge does not significantly influence their smoking behavior, it is noted that in the case of teachers, smoking reason (p<0.05), attitudes (p<0.001) and knowledge (p<0.05) strongly influence their smoking behavior. 8. There is a significantly correlation among the smoking reasons, attitudes and knowledge between students and parents(p<0.001). As for the correlationship between regularity and smoking amount and other influencing factors, the daily smoking amount is in proportion to depth of inhalation and duration of smoking, negative attitudes of parents unstability of family, dissatisfaction of family members, juvenile delinquency, strong smoking reasons and positive attitudes towards smoking. 9. In the case of daily smokers depth of inhalation is significantly correlated to the duration of smoking, juvenile delinquency, acceptability of parents, dissatisfaction of family members and smoking reasons. The duration of smoking motives is significantly correlated to juvenile delinquency, high acceptability of parents, strong smoking motive and positive attitudes toward smoking. 10. It is noted that 40% of parents and 30% of teachers do not recognize the significant correlationship between and the relative influencing factors mentioned above.