• ETRI Journal
• /
• v.46 no.2
• /
• pp.307-322
• /
• 2024

Most object detection methods use a horizontal bounding box that causes problems between adjacent objects with arbitrary directions, resulting in misaligned detection. Hence, the horizontal anchor should be replaced by a rotating anchor to determine oriented bounding boxes. A two-stage process of delineating a horizontal bounding box and then converting it into an oriented bounding box is inefficient. To improve detection, a box-boundary-aware vector can be estimated based on a convolutional neural network. Specifically, we propose a ResNeXt101 encoder to overcome the weaknesses of the conventional ResNet, which is less effective as the network depth and complexity increase. Owing to the cardinality of using a homogeneous design and multi-branch architecture with few hyperparameters, ResNeXt captures better information than ResNet. Experimental results demonstrate more accurate and faster oriented object detection of our proposal compared with a baseline, achieving a mean average precision of 89.41% and inference rate of 23.67 fps.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.16 no.1
• /
• pp.16-37
• /
• 2022

Accurate liver segment segmentation based on radiological images is indispensable for the preoperative analysis of liver tumor resection surgery. However, most of the existing segmentation methods are not feasible to be used directly for this task due to the challenge of exact edge prediction with some tiny and slender vessels as its clinical segmentation criterion. To address this problem, we propose a novel deep learning based segmentation model, called Boundary-Aware Dual Attention Liver Segment Segmentation Model (BADA). This model can improve the segmentation accuracy of liver segments with enhancing the edges including the vessels serving as segment boundaries. In our model, the dual gated attention is proposed, which composes of a spatial attention module and a semantic attention module. The spatial attention module enhances the weights of key edge regions by concerning about the salient intensity changes, while the semantic attention amplifies the contribution of filters that can extract more discriminative feature information by weighting the significant convolution channels. Simultaneously, we build a dataset of liver segments including 59 clinic cases with dynamically contrast enhanced MRI(Magnetic Resonance Imaging) of portal vein stage, which annotated by several professional radiologists. Comparing with several state-of-the-art methods and baseline segmentation methods, we achieve the best results on this clinic liver segment segmentation dataset, where Mean Dice, Mean Sensitivity and Mean Positive Predicted Value reach 89.01%, 87.71% and 90.67%, respectively.

• Journal of Intelligence and Information Systems
• /
• v.21 no.1
• /
• pp.15-28
• /
• 2015

As personal devices and pervasive technologies for interacting with networked objects continue to proliferate, there is an unprecedented world of scattered pieces of contextualized information available. However, the explosive growth and variety of information ironically lead users and service providers to make poor decision. In this situation, recommender systems may be a valuable alternative for dealing with these information overload. But they failed to utilize various types of contextual information. In this study, we suggest a methodology for context-aware recommender systems based on the concept of contextual boundary. First, as we suggest contextual boundary-based profiling which reflects contextual data with proper interpretation and structure, we attempt to solve complexity problem in context-aware recommender systems. Second, in neighbor formation with contextual information, our methodology can be expected to solve sparsity and cold-start problem in traditional recommender systems. Finally, we suggest a methodology about context support score-based recommendation generation. Consequently, our methodology can be first step for expanding application of researches on recommender systems. Moreover, as we suggest a flexible model with consideration of new technological development, it will show high performance regardless of their domains. Therefore, we expect that marketers or service providers can easily adopt according to their technical support.

• International Journal of Internet, Broadcasting and Communication
• /
• v.15 no.2
• /
• pp.14-21
• /
• 2023

Recent information and communication technologies have already opened up new prospects for technology groups, especially in a knowledge-based society. A contemporary technological era, which can be stated as the hyper-connective Internet of Things surpassed the traditional service pattern and innovation pattern by conveying personalized, localized, and con-text-aware services close to different actors and users. The conventional boundary of the organization is disbanding as well as traditional innovation and research & development limits. This research article conducts a preliminary study about the hyper-connective Internet of Things technology portent with innovation 3.0 version based on an extended technological organization environment framework (E-TOEF). This article discusses the emergence of innovation 3.0 as a paradigm shift from a manufacturing paradigm to an actor-oriented paradigm. There is a need to shift from a manufacturing mindset to more user ergonomics and be aware of the potential of hyper-connective IoT on the revolution of innovation patterns to be more cooperative, open, and user-centered. Besides, this article would strain some conceptual approaches for the next-generation innovation paradigm known as "hyper-connective IoT" entitled innovation 3.0. This new innovation version goes beyond open innovation and undeniably clearly beyond closed innovation which was an earlier version.

• Proceedings of the Korean Institute of Interior Design Conference
• /
• 2006.11a
• /
• pp.219-222
• /
• 2006

Architecture is builded on the ground and affected environment which include shape of topography and situation. What this study saying is the analysis about the relation between the concept of structural geomorphology and the space modern architecture. As there are many issue about boundary of space and organic architecture, It is important what study about structural geomorphology In paper, we will aware that there are many similarities between architecture and topography. Notion of folded structure in structural geomorphology is connected with continuity or infinity. This is one of many example.

• The Journal of Korea Robotics Society
• /
• v.17 no.2
• /
• pp.133-141
• /
• 2022

This paper introduces model compression algorithms which make a deep neural network smaller and faster for embedded systems. The model compression algorithms can be largely categorized into pruning, quantization and knowledge distillation. In this study, gradual pruning, quantization aware training, and knowledge distillation which learns the activation boundary in the hidden layer of the teacher neural network are integrated. As a large deep neural network is compressed and accelerated by these algorithms, embedded computing boards can run the deep neural network much faster with less memory usage while preserving the reasonable accuracy. To evaluate the performance of the compressed neural networks, we evaluate the size, latency and accuracy of the deep neural network, DenseNet201, for image classification with CIFAR-10 dataset on the NVIDIA Jetson Xavier.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2015.05a
• /
• pp.93-96
• /
• 2015

In this paper, we introduce a fire detection system using context-aware sensor. In existing weather and based on vision sensor of fire detection system case, acquired image through sensor of camera is extracting features about fire range as processing to convert HSI(Hue, Saturation, Intensity) model HSI which is color space can have durability in illumination changes. However, in this case, until a fire occurs wide range of sensing a fire in a single camera sensor, it is difficult to detect the occurrence of a fire. Additionally, the fire detection in complex situations as well as difficult to separate continuous boundary is set for the required area is difficult. In this paper, we propose an algorithm for real-time by using a temperature sensor, humidity, Co2, the flame presence information acquired and comparing the data based on multiple conditions, analyze and determine the weighting according to fire it. In addition, it is possible to differential management to intensive fire detection is required zone dividing the state of fire.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.11 no.2
• /
• pp.422-429
• /
• 2007

Location awareness is one of the key functionalities to build an U-city. Recently, many of works of the location-aware systems are emerging to commercially apply to on-going large-scale apartment complex based on U-city. As dwellers or cars being attached with active tags are moving in the U-city complex, the active tags periodically broadcast their own identifiers and receivers fixed along the street or in building use those information to calculate location of them. There are several issues to be considered for such an environment. The first is that the number of active tags operating in the same region are large as much as tens of thousands, and the second is that the active tags should be alive without change of batteries more than a year, hence low power consumption is very important. In this paper we propose i) a new architecture for location-aware system considering such issues, ii) technical issues to implement it using active tags, and iii) a mathematical analytic model to investigate overall performance and verify it by comparing with actual experimental results. Through the analysis we can show the theoretical boundary of the lowest packet loss rate and the maximum number of tags with acceptable performance for the systems based on active tags. The results can be applied to practical design of location-based systems of U-City projects.

• Journal of the Institute of Electronics and Information Engineers
• /
• v.49 no.10
• /
• pp.54-59
• /
• 2012

Multiple antenna transmission and reception, whose principal merits are significant increase in spectral efficiency and/or reduction in error rate, lose much of their effectiveness in high levels of interference from other cells. Incorporating the other cell interference into advanced signal processing at transmitter and receiver is one of the key challenges for cell boundary users in cellular system. Since receiver can obtain exact knowledge of interference channels more easily than transmitter, an interference-aware multiple antenna receiver that can significantly attenuate interferences is considered. Based on the receiver, codebook-based coordinated precoding schemes are proposed. According to the level of cooperation, centralized and distributed schemes are proposed. We verified by the simulation results that even the distributed schemes, which have same amount of feedback and no cooperation between cells, have performance gain compared to the conventional non-coordinated scheme.

• Journal of the Korean Society of Fashion and Beauty
• /
• v.3 no.2 s.2
• /
• pp.55-64
• /
• 2005

The purpose of this study is to investigate consumers' knowledge and preference towards functional cosmetics. Through the beauty advisors' surveys, their own selling styles and consumer behaviors recognized by beauty advisors were analyzed. It was intended to help extend and strengthen the functional cosmetic market which has continued to grow rapidly since the approved goods under cosmetic law in 2001. For this study, the data was collected through questionnaires the professional consumer counselors confirmed from Korean women over the age. of twenty old living in the Seoul and Kyoungki areas. After pre-research was implemented on 45 women, 328 samples were analyzed as final samples. In addition,46 samples, which were collected through the questionnaires from beauty advisors were analyzed. Samples were analyzed by frequency, percentage, T-test, ANOVA using the SPSS program. The results of study were as follows: First consumers recognized whether functional cosmetics or not. According to the beauty advisor's surveys, consumers regard the functional cosmetic boundary as being wider than real functional cosmetic boundary according to cosmetic law. So, there is a gap between consumers' opinions and real law. Second, regarding the purchasing channels, the largest channel is the cosmetic store. As far as consumers are concerned the most important factor when buying cosmetics, is the suitability of their own skin types. The second factor is product quality and the third factor is price. Functional cosmetics non-experienced group are more concerned with price compared to experienced group. Related to purchasing products, functional cosmetics experienced group buy set products compared to non-experienced group buy one product. Third, the ultraviolet filter cosmetics portion is the largest in the functional cosmetics market the second largest portion is bleaching cosmetics and the third largest portion is the anti-senility cosmetics. However, Most preferred by consumers is the anti-senility cosmetics. Moreover, preference for ultraviolet filter cosmetics is the least. Finally, the level of satisfaction for functional cosmetics is high and dissatisfaction is low. Consumers feel that beauty advisors are simply pushing high priced products without recognizing the consumers' real needs. In conclusion, to develop the functional cosmetic market continually in the future, it needs to extend various products and advertise them until consumers are more aware.