• Journal of the Institute of Electronics and Information Engineers
• /
• v.51 no.7
• /
• pp.157-168
• /
• 2014

We propose an novel artificial visual attention model that is capable of automatic detection and segmentation of saliency region on natural images in this paper. The proposed model is based on human visual perceptions in biological vision and contains there are main contributions. Firstly, we propose a novel framework of artificial visual attention model based on the opponent process theory using intensity and color features, and an entropy filter is designed to perceive salient regions considering the amount of information from intensity and color feature channels. The entropy filter is able to detect and segment salient regions in high segmentation accuracy and precision. Lastly, we also propose an adaptive combination method to generate a final saliency map. This method estimates scores about intensity and color conspicuous maps from each perception model and combines the conspicuous maps with weight derived from scores. In evaluation of saliency map by ROC analysis, the AUC of proposed model as 0.9256 approximately improved 15% whereas the AUC of previous state-of-the-art models as 0.7824. And in evaluation of salient region segmentation, the F-beta of proposed model as 0.7325 approximately improved 22% whereas the F-beta of previous state-of-the-art models.

• Journal of Intelligence and Information Systems
• /
• v.9 no.3
• /
• pp.23-47
• /
• 2003

Traditional studies for customer relationship management (CRM) generally focus on static CRM in a specific time frame. The static CRM and customer behavior knowledge derived could help marketers to redirect marketing resources fur profit gain at that given point in time. However, as time goes, the static knowledge becomes obsolete. Therefore, application of CRM to an online retailer should be done dynamically in time. Customer-based analysis should observe the past purchase behavior of customers to understand their current and likely future purchase patterns in consumer markets, and to divide a market into distinct subsets of customers, any of which may conceivably be selected as a market target to be reached with a distinct marketing mix. Though the concept of buying-behavior-based CRM was advanced several decades ago, virtually little application of the dynamic CRM has been reported to date. In this paper, we propose a dynamic CRM model utilizing data mining and a Monitoring Agent System (MAS) to extract longitudinal knowledge from the customer data and to analyze customer behavior patterns over time for the Internet retailer. The proposed model includes an extensive analysis about a customer career path that observes behaviors of segment shifts of each customer: prediction of customer careers, identification of dominant career paths that most customers show and their managerial implications, and about the evolution of customer segments over time. furthermore, we show that dynamic CRM could be useful for solving several managerial problems which any retailers may face.

• Journal of Korea Spatial Information System Society
• /
• v.10 no.3
• /
• pp.1-18
• /
• 2008

Because we can usually get many information through analyzing trajectories of moving objects on spatial networks, efficient trajectory index structures are required to achieve good retrieval performance on their trajectories. However, there has been little research on trajectory index structures for spatial networks such as FNR-tree and MON-tree. Also, because FNR-tree and MON-tree store the segment unit of moving objects, they can't support the trajectory of whole moving objects. In this paper, we propose an efficient trajectory index structures based on signatures on a spatial network, named SigMO-Tree. For this, we divide moving object data into spatial and temporal attributes, and design an index structure which supports not only range query but trajectory query by preserving the whole trajectory of moving objects. In addition, we divide user queries into trajectory query based on spatio-temporal area and similar-tralectory query, and propose query processing algorithms to support them. The algorithm uses a signature file in order to retrieve candidate trajectories efficiently Finally, we show from our performance analysis that our trajectory index structure outperforms the existing index structures like FNR-Tree and MON-Tree.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.26 no.6
• /
• pp.471-476
• /
• 2016

In this paper, recognition system for continuous human action is explained by using motion history image and histogram of oriented gradient with spotter model based on depth information, and the spotter model which performs action spotting is proposed to improve recognition performance in the recognition system. The steps of this system are composed of pre-processing, human action and spotter modeling and continuous human action recognition. In pre-processing process, Depth-MHI-HOG is used to extract space-time template-based features after image segmentation, and human action and spotter modeling generates sequence by using the extracted feature. Human action models which are appropriate for each of defined action and a proposed spotter model are created by using these generated sequences and the hidden markov model. Continuous human action recognition performs action spotting to segment meaningful action and meaningless action by the spotter model in continuous action sequence, and continuously recognizes human action comparing probability values of model for meaningful action sequence. Experimental results demonstrate that the proposed model efficiently improves recognition performance in continuous action recognition system.

• Journal of Internet of Things and Convergence
• /
• v.8 no.3
• /
• pp.29-40
• /
• 2022

The aim of this study was to analyze research trends of social service quality from 2007 to 2020 based on text mining and topic modeling. Our focus was to provide foundational materials for social service improvement by discovering the latent meaning of relevant research papers. We collected 97 scholarly articles on social service, social welfare service, and quality from RISS, and implemented two segments of text mining analysis. Our results showed that the first section included 38 papers and the second 59, indicating 6.9 articles annually. Word frequency results demonstrated that the common keywords of both sections were 'service', 'quality', 'social service', 'satisfaction', 'users', 'quality control', 'reuse', 'policy', 'voucher', etc. TF-IDF suggested that 'social service', 'satisfaction', 'users', 'customer satisfaction', 'revisiting', 'voucher', 'quality', 'assisted living facility', 'quality control', 'community service investment business', etc., were represented in both categories. Lastly, topic modeling analysis revealed that the first segment displayed 'types of care services', 'service costs', 'reuse', 'users based', and 'job creation', whereas the second presented 'service quality', 'public value', 'management system of human resources', 'service provision system', and 'service satisfaction'. Future directions of social service quality were discussed based on the results.

• KIPS Transactions on Computer and Communication Systems
• /
• v.12 no.3
• /
• pp.119-126
• /
• 2023

Sarcopenia is not well known enough to be classified as a disease in 2021 in Korea, but it is recognized as a social problem in developed countries that have entered an aging society. The diagnosis of sarcopenia follows the international standard guidelines presented by the European Working Group for Sarcopenia in Older People (EWGSOP) and the d Asian Working Group for Sarcopenia (AWGS). Recently, it is recommended to evaluate muscle function by using physical performance evaluation, walking speed measurement, and standing test in addition to absolute muscle mass as a diagnostic method. As a representative method for measuring muscle mass, the body composition analysis method using DEXA has been formally implemented in clinical practice. In addition, various studies for measuring muscle mass using abdominal images of MRI or CT are being actively conducted. In this paper, we develop an AI image segmentation model based on abdominal images of CT with a relatively short imaging time for the diagnosis of sarcopenia and describe the multicenter validation. We developed an artificial intelligence model using U-Net that can automatically segment muscle, subcutaneous fat, and visceral fat by selecting the L3 region from the CT image. Also, to evaluate the performance of the model, internal verification was performed by calculating the intersection over union (IOU) of the partitioned area, and the results of external verification using data from other hospitals are shown. Based on the verification results, we tried to review and supplement the problems and solutions.

• Journal of the Society of Disaster Information
• /
• v.20 no.2
• /
• pp.329-338
• /
• 2024

Purpose: This study verified the safety of the improved box-type girder behavior by comparing and evaluating the bending behavior results of a full-scale specimen based on the analytical behavior of the splice element PSC U-shaped girder with integrated tensioning systems. Method: Based on the results of the service and strength limit state design using the bridge design standard(limit state design method), the applied load of a 40m full-scale specimen was calculated and a static loading experiment using the four-point loading method was performed. Result: When the design load, crack load, and ultimate load were applied, the specimen deflection occurred at 97.1%, 98.5%, and 79.0% of the analytical deflection value. When the design load, crack load, and ultimate load were applied, the crack gauge was measured at 0.009~0.035mm, 0.014~0.050mm, and 6.383~5.522mm at each connection. Conclusion: The specimen behaved linear-elastically until the crack load was applied, and after cracks occurred, it showed strainhardening up to the ultimate load, and it was confirmed that the resistance of bending behavior was clearly displayed against the applied load. The cracks in the dry joints were less than 25% of grade B based on the evaluation of facility condition standard. The final residual deformation after removing the ultimate load was 0.114mm, confirming the stable behavior of the segment connection.

• The Journal of the Acoustical Society of Korea
• /
• v.28 no.6
• /
• pp.572-577
• /
• 2009

Large corpus-based concatenating Text-to-Speech (TTS) systems can generate natural synthetic speech without additional signal processing. To prune the redundant speech segments in a large speech segment DB, we can utilize a decision-tree based triphone clustering algorithm widely used in speech recognition area. But, the conventional methods have problems in representing the acoustic transitional characteristics of the phones and in applying context questions with hierarchic priority. In this paper, we propose a new clustering algorithm to downsize the speech DB. Firstly, three 13th order MFCC vectors from first, medial, and final frame of a phone are combined into a 39 dimensional vector to represent the transitional characteristics of a phone. And then the hierarchically grouped three question sets are used to construct the triphone trees. For the performance test, we used DTW algorithm to calculate the acoustic similarity between the target triphone and the triphone from the tree search result. Experimental results show that the proposed method can reduce the size of speech DB by 23% and select better phones with higher acoustic similarity. Therefore the proposed method can be applied to make a small sized TTS.

• Journal of Intelligence and Information Systems
• /
• v.19 no.4
• /
• pp.39-53
• /
• 2013

The emergence of the internet technology and SNS has increased the information flow and has changed the way people to communicate from one-way to two-way communication. Users not only consume and share the information, they also can create and share it among their friends across the social network service. It also changes the Social Media behavior to become one of the most important communication tools which also includes Social TV. Social TV is a form which people can watch a TV program and at the same share any information or its content with friends through Social media. Social News is getting popular and also known as a Participatory Social Media. It creates influences on user interest through Internet to represent society issues and creates news credibility based on user's reputation. However, the conventional platforms in news services only focus on the news recommendation domain. Recent development in SNS has changed this landscape to allow user to share and disseminate the news. Conventional platform does not provide any special way for news to be share. Currently, Social News Service only allows user to access the entire news. Nonetheless, they cannot access partial of the contents which related to users interest. For example user only have interested to a partial of the news and share the content, it is still hard for them to do so. In worst cases users might understand the news in different context. To solve this, Social News Service must provide a method to provide additional information. For example, Yovisto known as an academic video searching service provided time dependent metadata from the video. User can search and watch partial of video content according to time dependent metadata. They also can share content with a friend in social media. Yovisto applies a method to divide or synchronize a video based whenever the slides presentation is changed to another page. However, we are not able to employs this method on news video since the news video is not incorporating with any power point slides presentation. Segmentation method is required to separate the news video and to creating time dependent metadata. In this work, In this paper, a time dependent metadata-based framework is proposed to segment news contents and to provide time dependent metadata so that user can use context information to communicate with their friends. The transcript of the news is divided by using the proposed story segmentation method. We provide a tag to represent the entire content of the news. And provide the sub tag to indicate the segmented news which includes the starting time of the news. The time dependent metadata helps user to track the news information. It also allows them to leave a comment on each segment of the news. User also may share the news based on time metadata as segmented news or as a whole. Therefore, it helps the user to understand the shared news. To demonstrate the performance, we evaluate the story segmentation accuracy and also the tag generation. For this purpose, we measured accuracy of the story segmentation through semantic similarity and compared to the benchmark algorithm. Experimental results show that the proposed method outperforms benchmark algorithms in terms of the accuracy of story segmentation. It is important to note that sub tag accuracy is the most important as a part of the proposed framework to share the specific news context with others. To extract a more accurate sub tags, we have created stop word list that is not related to the content of the news such as name of the anchor or reporter. And we applied to framework. We have analyzed the accuracy of tags and sub tags which represent the context of news. From the analysis, it seems that proposed framework is helpful to users for sharing their opinions with context information in Social media and Social news.

• The Korean Journal of Nuclear Medicine Technology
• /
• v.16 no.2
• /
• pp.68-80
• /
• 2012

Purpose : More recently, combined PET/MR scanners have been developed in which the MR data can be used for both anatometabolic image formation and attenuation correction of the PET data. For quantitative PET information, correction of tissue photon attenuation is mandatory. The attenuation map is obtained from the CT scan in the PET/CT. In the case of PET/MR, the attenuation map can be calculated from the MR image. The purpose of this study was to assess the quantitative differences between MR-based and CT-based attenuation corrected PET images. Materials and Methods : Using the uniform cylinder phantom of distilled water which has 199.8 MBq of $^{18}F$-FDG put into the phantom, we studied the effect of MR-based and CT-based attenuation corrected PET images, of the PET-CT using time of flight (TOF) and non-TOF iterative reconstruction. The images were acquired from 60 minutes at 15-minute intervals. Region of interests were drawn over 70% from the center of the image, and the Scanners' analysis software tools calculated both maximum and mean SUV. These data were analyzed by one way-anova test and Bland-Altman analysis. MR images are segmented into three classes(not including bone), and each class is assigned to each region based on the expected average attenuation of each region. For clinical diagnostic purpose, PET/MR and PET/CT images were acquired in 23 patients (Ingenuity TF PET/MR, Gemini TF64). PET/CT scans were performed approximately 33.8 minutes after the beginnig of the PET/MR scans. Region of interests were drawn over 9 regions of interest(lung, liver, spleen, bone), and the Scanners' analysis software tools calculated both maximum and mean SUV. The SUVs from 9 regions of interest in MR-based PET images and in CT-based PET images were compared. These data were analyzed by paired t test and Bland-Altman analysis. Results : In phantom study, MR-based attenuation corrected PET images generally showed slightly lower -0.36~-0.15 SUVs than CT-based attenuation corrected PET images (p<0.05). In clinical study, MR-based attenuation corrected PET images generally showed slightly lower SUVs than CT-based attenuation corrected PET images (excepting left middle lung and transverse Lumbar) (p<0.05). And percent differences were -8.01.79% lower for the PET/MR images than for the PET/CT images. (excepting lung) Based on the Bland-Altman method, the agreement between the two methods was considered good. Conclusion : PET/MR confirms generally lower SUVs than PET/CT. But, there were no difference in the clinical interpretations made by the quantitative comparisons with both type of attenuation map.