• Journal of the Korea Society of Computer and Information
• /
• v.3 no.4
• /
• pp.10-18
• /
• 1998

Video information is an important component of multimedia system such as Digital Library. World-Wide Web(WWW) and Video-On-Demand(VOD) service system. It can support various types of information because of including audio-visual, spatial-temporal and semantics information. In addition, it requires the ability of retrieving the specific scene of video instead of entire retrieval of video document. Therefore, so as to support a variety of retrieval, this paper models metadata using video document structure information that consists of hierarchical structure, and designs database schema that can manipulate video document.

• Journal of the Korea institute for structural maintenance and inspection
• /
• v.2 no.2
• /
• pp.151-161
• /
• 1998

In this study, an attempt is made to apply the concept of fuzzy-bayesian theory to the integrity assessment of RC highway bridge, and uncertainty states are represented in terms of fuzzy sets which define several linguistic variables such as "very good", "good", "average", "poor", "very poor", etc. Especially, the concept of fuzzy conditional probability aids to derive a new reliability analysis which includes the subjective assessment of engineers without introducing any additional correction factors. The fuzzy concept are also used as reliability indexes for the condition assessment based on the proposed models, the proposed fuzzy theory-based approach with the results of visual inspection and extensive field load tests are applied to the integrity assessment of a new RC highway bridge, namely, Jichok bridge.

• Journal of the Korea Society for Simulation
• /
• v.11 no.4
• /
• pp.1-14
• /
• 2002

This paper presents bounding space(BS) management for real-time visual simulation when using GK-DEVS models. Since GK-DEVS, extended from DEVS formalism, has information of 3D geometry and 3D hierarchical structure, we employs three types of bounding spaces: BS of its own shape, BS of its children GK-DEVS, and total BS. In addition to next-event scheduling functionality of previous GK-Simulator, its abstract simulation algorithms is extended to manage the three types of BSs so that BSs can be utilized in the rendering process of a renderer, so called GK-Renderer, We have implemented the method and evaluated it with an automated manufacturing system. In the case study, the proposed BSs management method showed about 2 times improvement in terms of rendering process speed.

• ETRI Journal
• /
• v.41 no.4
• /
• pp.483-493
• /
• 2019

Considering the high dimensions of video sequences, it is often challenging to acquire a sufficient dataset to train the tracking models. From this perspective, we propose to revisit the idea of hand-crafted feature learning to avoid such a requirement from a dataset. The proposed tracking approach is composed of two phases, detection and tracking, according to how severely the appearance of a target changes. The detection phase addresses severe and rapid variations by learning a new appearance model that classifies the pixels into foreground (or target) and background. We further combine the raw pixel features of the color intensity and spatial location with convolutional feature activations for robust target representation. The tracking phase tracks a target by searching for frame regions where the best pixel-level agreement to the model learned from the detection phase is achieved. Our two-phase approach results in efficient and accurate tracking, outperforming recent methods in various challenging cases of target appearance changes.

• IEMEK Journal of Embedded Systems and Applications
• /
• v.17 no.6
• /
• pp.347-353
• /
• 2022

In this paper, we propose a method that performs a text-video retrieval model by replacing video properties using captions. In general, the exisiting embedding-based models consist of both joint embedding space construction and the CNN-based video encoding process, which requires a lot of computation in the training as well as the inference process. To overcome this problem, we introduce a video-captioning module to replace the visual property of video with captions generated by the video-captioning module. To be specific, we adopt the caption generator that converts candidate videos into captions in the inference process, thereby enabling direct comparison between the text given as a query and candidate videos without joint embedding space. Through the experiment, the proposed model successfully reduces the amount of computation and inference time by skipping the visual processing process and joint embedding space construction on two benchmark dataset, MSR-VTT and VATEX.

• Journal of the Society of Naval Architects of Korea
• /
• v.57 no.3
• /
• pp.140-151
• /
• 2020

It is important to identify the location of ship hull blocks with exact block identification number when scheduling the shipbuilding process. The wrong information on the location and identification number of some hull block can cause low productivity by spending time to find where the exact hull block is. In order to solve this problem, it is necessary to equip the system to track the location of the blocks and to identify the identification numbers of the blocks automatically. There were a lot of researches of location tracking system for the hull blocks on the stockyard. However there has been no research to identify the hull blocks on the stockyard. This study compares the performance of 5 Convolutional Neural Network (CNN) models with multi-view image set on the classification of the hull blocks to identify the blocks on the stockyard. The CNN models are open algorithms of ImageNet Large-Scale Visual Recognition Competition (ILSVRC). Four scaled hull block models are used to acquire the images of ship hull blocks. Learning and transfer learning of the CNN models with original training data and augmented data of the original training data were done. 20 tests and predictions in consideration of five CNN models and four cases of training conditions are performed. In order to compare the classification performance of the CNN models, accuracy and average F1-Score from confusion matrix are adopted as the performance measures. As a result of the comparison, Resnet-152v2 model shows the highest accuracy and average F1-Score with full block prediction image set and with cropped block prediction image set.

• Journal of the Institute of Convergence Signal Processing
• /
• v.25 no.2
• /
• pp.113-118
• /
• 2024

Recently, advancements in artificial intelligence and computer graphics technology have led to the emergence of various virtual humans across multiple media such as movies, advertisements, broadcasts, games, and social networking services (SNS). In particular, in the advertising marketing sector centered around virtual influencers, virtual humans have already proven to be an important promotional tool for businesses in terms of time and cost efficiency. In Korea, the virtual influencer market is in its nascent stage, and both large corporations and startups are preparing to launch new services related to virtual influencers without clear boundaries. However, due to the lack of public disclosure of the development process, they face the situation of having to incur significant expenses. To address these requirements and challenges faced by businesses, this paper implements a photogrammetry-based facial capture system for creating realistic virtual humans and explores the use of these models and their application cases. The paper also examines an optimal workflow in terms of cost and quality through MetaHuman modeling based on Unreal Engine, which simplifies the complex CG work steps from facial capture to the actual animation process. Additionally, the paper introduces cases where virtual humans have been utilized in SNS marketing, such as on Instagram, and demonstrates the performance of the proposed workflow by comparing it with traditional CG work through an Unreal Engine-based workflow.

• Journal of Korean Society of Forest Science
• /
• v.86 no.1
• /
• pp.87-97
• /
• 1997

In this research, photo images of uncultivated marginal lands were simulated to visualize alternative land use patterns using image capture technology. Based on an original photos, 3 simulated images were created ; barren condition, aforested condition and shrub-covered condition. The simulated images were then used to evaluate respondents' visual preference(SBE value) and willingness to pay for the agricultural development tax as a hypothetical payment vehicle. The SBE values for barren condition are the lowest, as expected. When original condition is changed to forested or shrubbed, the SBE values are increased significantly. The logistic models for the willingness to pay for the various alternative land uses performed significantly, ${\rho}$ statistics for 6 models ranges from 0.3 to 0.4 and correct percentage for predicted probability are about 75%. Among independent variables, the amount of tax offered is the most influencing factor to predict the probability. Income also shows some relationship with no statistical significance. Other variables behave inconsistently in the model. When SBE and WTP are correlated, rather consistent trends can be observed. With the increase of SBE, WTP predicted by the model increases accordingly. It can be concluded that enhancement of scenic quality of the agricultural lands leads to increase of people's willingness to pay to support the rural environmental conservation.

• 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 the Korean Society of Environmental Restoration Technology
• /
• v.1 no.1
• /
• pp.28-44
• /
• 1998

This study was conducted to suggest the ecological restoration strategies for the disturbed landscapes by theoretical study. Especially, it is aimed to suggest three objectives for restoration by using native plants ; (1) prevention or reduction of wind and water erosion, (2) provision of food and cover for variety of animal species, (3) improvement of the visual or aesthetic quality of disturbed sites. The main results were summarized as follows. 1. Cooperation between restoration ecologists and restoration practitioners is needed to delineate a scientific approach to restoration ; (1) Being aware of published literature that describes similar work and/or establish general principles. (2) Preparedness to carry out proper experiments to test ideas. (3) Preparedness to monitor fundamental parameters in a restoration scheme. (4) Providing information about the behavior of species. (5) Publish results. 2. There are three models of succession in theories of plant succession. The tolerance and facilitation models were recommended to ecological restoration. The inhibition model applies in most secondary succession, but the actual species which are introduced first may inhibit the germination and growth of slower-growing species, or they may prevent the growth of other species whose propagules arrive later. 3. The objectives of erosion control, wildlife habitat provision, and visual quality improvement are not mutually exclusive. However, many revegetation practices in the past have emphasized one of these aspects at the expense of the others. 4. A native plant community can be the model of ecological restoration. By stylization/abstraction of native plant community, trying to learn the most essential characteristics of community types - environmental factors ; dominant, prevalent, and "visual essence" species composition - in order to use such information in restoration. 5. After developing mass/spaces plan, match plant communities to the mass/space plan. In utilizing community grouping, there needs aesthetic ability to understand design elements. 6. Several hydrophytes such as Pennisetum saccharifluous, Themeda triandra, Cirsium pendulum show relatively good germination rates. In case of mesophytes and xerophytes, Arundinella hina, Artemisia princeps, Oenothera odorata and legumes seem to have quick-germinating abilities at barren sites. Pinus thunbergii, Rhus chinensis, Evodia daniellii, Alnus firma and Albizzia julibrissin can be considered as "late succession" woody plants because they show low germinating rates and slow growing habitat. 7. The seeds used for restoration should be collected within a certain radius of where it will be planted. Consideration in genetic issues in the collection and use of germplasm can increase the odds for successful restoration efforts. 8. The useful model in the "drift" pattern occurs so abundantly in naturally evolving landscapes. As one species diminishes in density, a second or third species are increased. Thus, dynamic interactions between species are created. Careful using of "drift phenomenon" in planting was recommended. 9. Virtually no stand of vegetation today is immune from the introduction and/or spreading of exotic species. Therefore, the perpetuation of a restored sites requires conscious monitoring and management. Thus, management would be the most important process in ecological restoration. 10. In order to keep the sites "natural", alternative management strategies would be applied in management ; Atenative management strategies are decelerating successional process, accelerating successional processes or modifying species composition. As management tools, mechanical, chemical, biological or environmental factor manipulation, e.g., fire could be used.