• Journal of KIISE:Databases
• /
• v.32 no.4
• /
• pp.362-371
• /
• 2005

Similarity indexing and searching are well known to be difficult in high-dimensional applications such as multimedia databases. Especially, they become more difficult when multiple features have to be indexed together. In this paper, we propose a novel indexing method called the GB-index that is designed to efficiently handle complex similarity queries as well as relevance feedback in high-dimensional image databases. In order to provide the flexibility in controlling multiple features and query objects, the GB-index treats each dimension independently The efficiency of the GB-index is realized by specialized bitmap indexing that represents all objects in a database as a set of bitmaps. Main contributions of the GB-index are three-fold: (1) It provides a novel way to index high-dimensional data; (2) It efficiently handles complex similarity queries; and (3) Disjunctive queries driven by relevance feedback are efficiently treated. Empirical results demonstrate that the GB-index achieves great speedups over the sequential scan and the VA-file.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.6 no.2
• /
• pp.271-281
• /
• 2002

Currently, most web designers guarante high quality using bitmap graphics as fixed font size, but that has defects about file size and flexibility. Especially, to provide high quality of banner and advertise characters, after you should use a bitmap edit program, and then we should follow the method we add that program to HTML documents as bitmap data. In this study, as I show a couple of new tags in front of HTML documents, I show methods which can be presented diverse effects. When text information are stored, because we print out a screen with simple control points and outside information, it can be possible for us to express the same quality of Hangul characters like printed documents in a web browser. Regardless of the second class of platform, we can make it possible the character expression with exact character expressions and diverse effects.

• The Journal of the Korea institute of electronic communication sciences
• /
• v.12 no.1
• /
• pp.139-146
• /
• 2017

The interest about Artificial Intelligence through the AlphaGo Match in Korea has been increasing rapidly. So far, very little has been done in Artificial Intelligence. The aim of this paper is to reveal technology convergence and to assess the country profile in the field of artificial intelligence(AI). Technology convergence map was created after extracting USPTO patent grants and Web of Science data and generating matrics in the field of AI. Several Indicators were obtained by extracting and calculating SCOPUS Data that KISTI has. According to USPTO patent grants, it shows that AI technology has a strong relationship with several sectors such as cost/price determination, image analysis, and surgery, etc. Also, AI has a active convergence with some fields of Electrical and Electronic Engineering, BioTechnologies, and Medicine etc. According to country profile analysis, Korea reaches a global average growth index. However, in terms of specialization index (SI) and average of relative citations (ARC), there is a large gap between Korea and research leading countries.

• KIPS Transactions on Software and Data Engineering
• /
• v.10 no.9
• /
• pp.349-358
• /
• 2021

Recently, a method of finding the attention area or localization area for an object of an image using CAM (Class Activation Map)[1] has been variously carried out as a study of WSOL (Weakly Supervised Object Localization). The attention area extraction from the object heat map using CAM has a disadvantage in that it cannot find the entire area of the object by focusing mainly on the part where the features are most concentrated in the object. To improve this, using CAM and Selective Search[6] together, we first expand the attention area in the heat map, and a Gaussian smoothing is applied to the extended area to generate retraining data. Finally we train the data to expand the attention area of the objects. The proposed method requires retraining only once, and the search time to find an localization area is greatly reduced since the selective search is not needed in this stage. Through the experiment, the attention area was expanded from the existing CAM heat maps, and in the calculation of IOU (Intersection of Union) with the ground truth for the bounding box of the expanded attention area, about 58% was improved compared to the existing CAM.

• Journal of the Korea Convergence Society
• /
• v.12 no.10
• /
• pp.55-61
• /
• 2021

In this paper, we proposed a DeepLabv3+ based encoder-decoder model utilizing an attention mechanism for precise semantic segmentation. The DeepLabv3+ is a semantic segmentation method based on deep learning and is mainly used in applications such as autonomous vehicles, and infrared image analysis. In the conventional DeepLabv3+, there is little use of the encoder's intermediate feature map in the decoder part, resulting in loss in restoration process. Such restoration loss causes a problem of reducing segmentation accuracy. Therefore, the proposed method firstly minimized the restoration loss by additionally using one intermediate feature map. Furthermore, we fused hierarchically from small feature map in order to effectively utilize this. Finally, we applied an attention mechanism to the decoder to maximize the decoder's ability to converge intermediate feature maps. We evaluated the proposed method on the Cityscapes dataset, which is commonly used for street scene image segmentation research. Experiment results showed that our proposed method improved segmentation results compared to the conventional DeepLabv3+. The proposed method can be used in applications that require high accuracy.

• Journal of the Korea Society of Computer and Information
• /
• v.29 no.10
• /
• pp.35-45
• /
• 2024

In this paper, we propose Multi-Scale Feature Knowledge Distillation for 3D Object Detection (M3KD), which extracting knowledge from the teacher model, and transfer to the student model consider with multi-scale feature map. To achieve this, we minimize L2 loss between feature maps at each pyramid level of the student model with the correspond teacher model so student model can mimic the teacher model backbone information which improves the overall accuracy of the student model. We apply the class logits knowledge distillation used in the image classification task, by allowing student model mimic the classification logits of the teacher model, to guide the student model to improve the detection accuracy. In KITTI (Karlsruhe Institute of Technology and Toyota Technological Institute) dataset, our M3KD (Multi-Scale Feature Knowledge Distillation for 3D Object Detection) student model achieves 30% inference speed improvement compared to the teacher model. Additionally, our method achieved an average improvement of 1.08% in 3D mean Average Precision (mAP) across all classes and difficulty levels compared to the baseline student model. Furthermore, when integrated with the latest knowledge distillation methods such as PKD and SemCKD, our approach achieved an additional 0.42% and 0.52% improvement in 3D mAP, respectively, further enhancing performance.

• Journal of the Korean Institute of Landscape Architecture
• /
• v.49 no.4
• /
• pp.15-29
• /
• 2021

The study aims to grasp the perception and appraisal of urban park users through text analysis. This study used Google review data provided by Google Maps. Google Maps Review is an online review platform that provides information evaluating locations through social media and provides an understanding of locations from the perspective of general reviewers and regional guides who are registered as members of Google Maps. The study determined if the Google Maps Reviews were useful for extracting meaningful information about the user perceptions and appraisals for parks management plans. The study chose three urban parks in Seoul, South Korea; Seoul Forest, Boramae Park, and Olympic Park. Review data for each of these three parks were collected via web crawling using Python. Through text analysis, the keywords and network structure characteristics for each park were analyzed. The text was analyzed, as were park ratings, and the analysis compared the reviews of residents and foreign tourists. The common keywords found in the review comments for the three parks were "walking", "bicycle", "rest" and "picnic" for activities, "family", "child" and "dogs" for accompanying types, and "playground" and "walking trail" for park facilities. Looking at the characteristics of each park, Seoul Forest shows many outdoor activities based on nature, while the lack of parking spaces and congestion on weekends negatively impacted users. Boramae Park has the appearance of a city park, with various facilities providing numerous activities, but reviewers often cited the park's complexity and the negative aspects in terms of dog walking groups. At Olympic Park, large-scale complex facilities and cultural events were frequently mentioned, emphasizing its entertainment functions. Google Maps Review can function as useful data to identify parks' overall users' experiences and general feelings. Compared to data from other social media sites, Google Maps Review's data provides ratings and understanding factors, including user satisfaction and dissatisfaction.

• Journal of the Korean Data and Information Science Society
• /
• v.21 no.1
• /
• pp.11-20
• /
• 2010

The purpose of this study is to analyze the images of the Military Academy, Naval Academy, Air Force Academy, Police Academy, and Armed Forces Nursing Academy using multidimensional scaling method. For this research, we surveyed 363 applicants to special purpose academies including Military Academies and Police College. The study showed that the Military, Naval, and Air Force Academies had stronger image than the Police Academy in the area of physical strength, tradition, and fellowship between senior and junior. On the other hand, the Police Academy had better image in the area of social activity and applicant's academic achievement. The Military Academy had been evaluated the best school among the three Academies in the area of applicant's academic achievement, educational environment, faculty, tradition, and fellowship between senior and junior.

• Journal of Digital Contents Society
• /
• v.18 no.8
• /
• pp.1635-1640
• /
• 2017

In this paper, a hierarchical image tree model for weather classification is defined in a single outdoor image, and a weather classification algorithm using image intensity and k-mean segmentation image is proposed. In the first level of the hierarchical image tree model, the indoor and outdoor images are distinguished. Whether the outdoor image is daytime, night, or sunrise/sunset image is judged using the intensity and the k-means segmentation image at the second level. In the last level, if it is classified as daytime image at the second level, it is finally estimated whether it is sunny or foggy image based on edge map and fog rate. Some experiments are conducted so as to verify the weather classification, and as a result, the proposed method shows that weather features are effectively detected in a given image.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.11 no.8
• /
• pp.1442-1449
• /
• 2007

Shadows are important elements for realistic rendering of the 3D scene. We cannot recognize the distance of objects in the 3D scene without shadows. Two methods, image-based medthods and object-based methods, are largely used for the rendering of shadows. Object based methods can generate accurate shadow boundaries. However, it cannot be used to generate the realtime shadows because the time complexity defends on the complexity of the 3D scene. Image based methods which are techniques to generate shadows are widely used because of fast calculation time. However, this algorithm has aliasing problems. PCF is a method to solve the aliasing problem. Using PCF technique, antialiased shadow boundaries can be generated. However, PCF with large filter size requires more time to calculate antialiased shadow boundaries. This paper proposes an improved PCF technique which generates antialiased shadow boundaries similar to that of PCF. Compared with PCF, this technique can generate antialiased shadows in less time.