• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.17 no.3
• /
• pp.861-880
• /
• 2023

Semantic segmentation of road scene is the key technology of autonomous driving, and the improvement of convolutional neural network architecture promotes the improvement of model segmentation performance. The existing convolutional neural network has the simplification of learning knowledge and the complexity of the model. To address this issue, we proposed a road scene semantic segmentation algorithm based on multi-task collaborative learning. Firstly, a depthwise separable convolution atrous spatial pyramid pooling is proposed to reduce model complexity. Secondly, a collaborative learning framework is proposed involved with saliency detection, and the joint loss function is defined using homoscedastic uncertainty to meet the new learning model. Experiments are conducted on the road and nature scenes datasets. The proposed method achieves 70.94% and 64.90% mIoU on Cityscapes and PASCAL VOC 2012 datasets, respectively. Qualitatively, Compared to methods with excellent performance, the method proposed in this paper has significant advantages in the segmentation of fine targets and boundaries.

• Trans-
• /
• v.5
• /
• pp.65-82
• /
• 2018

What distinguishes Impression Liu Sanjie, directed by Zhang yimou from other stage performances lies in its use of space. Firstly, it breaks the regular way of performing in a limited theater space but instead, impression Liu sanjie bases the stage activities on the real nature scenery. Secondly, it promotes the local culture by combining the unique color effect of the particular nature scenery with its use of the scenery's natural space, with an aesthetic precision. In another word, Impression Liu Sanjie is not a simple performance set on the natural scenery but more of a promotion to the local culture, through the concept of "environment friend" between human and nature, which is brought by mixing the unique color of natural scenery and its natural space. This research analyzes the Impression Liu Sanjie's use of space basing on Proxemics by Edward T Hall. Impression Liu Sanjie adapts its character from a folk tale, employs the form of natural scenery performance, and at the same time to the largest extent throughout its creating process respects the local culture and its natural space. As a result, the performance presents the idea of "environmentally friendly".

• Journal of Korea Multimedia Society
• /
• v.11 no.5
• /
• pp.597-609
• /
• 2008

Due to the recursive updating nature of background model, previous background modeling methods are often perturbed by crowd scenes where foreground pixels occurs more frequently than background pixels. To resolve this problem, an adaptive background modeling method, which is based on the well-known Gaussian mixture background model, is proposed. In the proposed method, the learning rate of background model is adaptively adjusted with respect to the crowdedness of the scene. Consequently, the learning process is suppressed in crowded scene to maintain proper background model. Experiments on real dataset revealed that the proposed method could perform background subtraction effectively even in crowd situation while the performance is almost the same to the previous method in normal scenes. Also, the F-measure was increased by 5-10% compared to the previous background modeling methods in the video of crowded situations.

• Proceedings of the Korean Institute Of Construction Engineering and Management
• /
• 2007.11a
• /
• pp.909-914
• /
• 2007

Today's dynamic nature of the construction environment requires management systems to be active enough to take real-time decisions. For real-time decision making, effective 3D spatial information acquisition is imperative. Various 3D data acquisition technologies are being developed and tested for 3D spatial information acquisition and its use for wide range of areas in the construction industry, however, there are shortcomings in these technologies. The major problems are long processing time and high cost which make current technologies impossible to be used for real-time applications. Laser-based Flash LADAR that illuminates the entire scene with diffuse laser light is comparatively fast and cost effective, therefore it is well suited for 3D spatial modeling of dynamic environment on a construction site. This paper presents experimental results to evaluate the performance of flash LADAR and discuss issues of applicability of Flash LADAR to 3D spatial modeling on a construction site.

• Journal of Intelligence and Information Systems
• /
• v.19 no.4
• /
• pp.81-96
• /
• 2013

Video information retrieval has become a very important issue because of the explosive increase in video data from Web content development. Meanwhile, content-based video analysis using visual features has been the main source for video information retrieval and browsing. Content in video can be represented with content-based analysis techniques, which can extract various features from audio-visual data such as frames, shots, colors, texture, or shape. Moreover, similarity between videos can be measured through content-based analysis. However, a movie that is one of typical types of video data is organized by story as well as audio-visual data. This causes a semantic gap between significant information recognized by people and information resulting from content-based analysis, when content-based video analysis using only audio-visual data of low level is applied to information retrieval of movie. The reason for this semantic gap is that the story line for a movie is high level information, with relationships in the content that changes as the movie progresses. Information retrieval related to the story line of a movie cannot be executed by only content-based analysis techniques. A formal model is needed, which can determine relationships among movie contents, or track meaning changes, in order to accurately retrieve the story information. Recently, story-based video analysis techniques have emerged using a social network concept for story information retrieval. These approaches represent a story by using the relationships between characters in a movie, but these approaches have problems. First, they do not express dynamic changes in relationships between characters according to story development. Second, they miss profound information, such as emotions indicating the identities and psychological states of the characters. Emotion is essential to understanding a character's motivation, conflict, and resolution. Third, they do not take account of events and background that contribute to the story. As a result, this paper reviews the importance and weaknesses of previous video analysis methods ranging from content-based approaches to story analysis based on social network. Also, we suggest necessary elements, such as character, background, and events, based on narrative structures introduced in the literature. We extract characters' emotional words from the script of the movie Pretty Woman by using the hierarchical attribute of WordNet, which is an extensive English thesaurus. WordNet offers relationships between words (e.g., synonyms, hypernyms, hyponyms, antonyms). We present a method to visualize the emotional pattern of a character over time. Second, a character's inner nature must be predetermined in order to model a character arc that can depict the character's growth and development. To this end, we analyze the amount of the character's dialogue in the script and track the character's inner nature using social network concepts, such as in-degree (incoming links) and out-degree (outgoing links). Additionally, we propose a method that can track a character's inner nature by tracing indices such as degree, in-degree, and out-degree of the character network in a movie through its progression. Finally, the spatial background where characters meet and where events take place is an important element in the story. We take advantage of the movie script to extracting significant spatial background and suggest a scene map describing spatial arrangements and distances in the movie. Important places where main characters first meet or where they stay during long periods of time can be extracted through this scene map. In view of the aforementioned three elements (character, event, background), we extract a variety of information related to the story and evaluate the performance of the proposed method. We can track story information extracted over time and detect a change in the character's emotion or inner nature, spatial movement, and conflicts and resolutions in the story.

• Journal of Korean Theatre Studies Association
• /
• no.48
• /
• pp.261-288
• /
• 2012

This paper attempts to research on the new language in the directing of Ara Kim. She was cranky on working on the stage to experiment with her own style since the 1980s and so opened a new dawn in modern Korean theatre. She leaded the Korean experimental theatre. The background of this experiment is her idea on theatre. And here, we have to look the subject that she setted for the work in Chuksan: Ritual Past, Ritual Present. To her, the theatre has the function of ritual and fest. The theatre suggests universal tragedy given to human as natural life force and has its own agenda to drive people to healing. For it, Ara Kim explores archetypal forms and languages before the fragmentation of genres of art. Her theatre shows the results of experiments in which such languages are recreated with modernized sensibilities. We here, for example by outdoor performance in Chuksan Human Lear, try to interpret the aesthetic principles that body out her ritual theatre. And what we looked at though, is the base of the 'complex-genre-music-theatre', the methode to 'compose' the stage elements and put it all together. The directing of Ara Kim has, in terms of the composition of the stage elements, much of the indisputable artistic value. Her theatre is, so to speak, theatre of image, and it is theatre of imagination that completed by the audience's imagination. Human Lear which has its own characteristic in image fragments, convert the original Lear into a simple tale. It serves as background of the modern ritual that shows the most basic human instincts. We meet in Human Lear a ritual tale with some list of image for the human instincts. The arrangement of image, the montage of scene shows the performance as a kind of artistic space. In Human Lear the space is the natural one. It centers around the arena stage. The objects installed in the space changes it into the laboratory for 'seeing' the happening. The spectators see the performance and at the same time see themselves in the nature laboratory. They see, and equally, they are visible objects. They see the performance and us in the space in which the performance takes place. That is what Ara Kim with her modern ritual really aims. That aim is to this days still in effect. It is a major driver of her experiments to extend the boundary of the theatre. The ritualistic site-specific performance in Akor Wat, Cambodia, A Song of Mandala is the latest great product from her experiments. On the other hand, she continues on her way to experiment with pure stage elements. The 'Station' series(Station of Water, The Station of Sand, The Station of Wind) she recently showed are the non-verbal performance with all the stage elements: movement, sound, body, light, colour, objects and so on.

• Journal of Korea Entertainment Industry Association
• /
• v.13 no.4
• /
• pp.165-172
• /
• 2019

Signs of the street and scenes of the nature(tree) popping up like a still screen in a monotonously repeated routine, as if waiting for something by chance, characters appearing and disappearing without special meaning, advise you to get away from cliché meetings and breakups while repeating it. The images of people returning to their places without presenting any conclusions highlights the coincidences and unexpectedness of everyday life, but are so familiar that it is easy to be overlooked, it brings us to experience the potential of everyday unfamiliarity in such moments. The actor's performance, which deviates from the practice of well-organized characters, is improvised (unplanned and uncoordinated) emphasized on the scene and expresses the everyday nature of his differentiated character freely like the reality as if he were locked up in a film structure but not confined to it, and as if he got used to the everyday life. Also the repeated words and actions of characters, and unpredictable changes of emotions, centered by the encounters and conversations of this film of Hong Sang-soo, everything that is presented in the background of everyday life in its form, let's us pay attention to the meaning and what to be found in the film. In addition, director Hong Sang-soo, who creates a story after selecting actors to reproduce this daily process in a realistic manner, sets new relationships between the characters and the actors, presenting a new method of realistic expression, through the actor who repeatedly appears in the film. This study analyzes the everyday characters and acting of the film based on director Hong Sang-soo's directing style.

• Korean Journal of Construction Engineering and Management
• /
• v.17 no.3
• /
• pp.116-124
• /
• 2016

Work efficiency of earth work which is one of the main works occurring in construction site mainly depends on the performance of individual operators of earth work equipment. Consequently, the skill of individual operators of earth work equipment can significantly affect overall construction schedules. Many invisible areas inevitably exist in construction site because of the nature of construction site where occlusions occur from structures being built, installed or moving equipment, moving workers, etc. The lack of visual information regarding tasks critically impedes the effective performance of operators of earth work equipment. AR (Augmented Reality) is a computer technology that superimposes virtual objects onto the real world scene. This characteristic of AR may address the lack of visual informations in earth work process, thus helping to improve the work efficiency of operators of earth work equipment. The purpose of this study is to present a task-visual information map that identifies visual informations required in tasks of earth work and which of the tasks are suitable for AR technology. This study focuses on visual informations in tasks of earth work with excavators. The map was created based on the investigations on the problems of each task of earth work with excavators and visual informations required to address the problems. Through the map, four visual informations were found to be suitable for AR technology to improve the work efficiency of excavator operators. Based on the findings of this study, AR systems for excavators can be developed more effectively.

• The Journal of the Korea Contents Association
• /
• v.12 no.12
• /
• pp.605-615
• /
• 2012

Dance studies inherently have a nature of convergence and integration. Meanwhile, dance studies have extended their realm by investigating dance phenomena from many perspectives which art theories cannot explain. However, the previous and current dance studies are inadequate to explain a confluence society which is characterized as techuim, Interaction, freedom and openness according to the digital revolution. Hence, a result of research trend in domestic dance studies, it is found that dance studies have been studied in four perspectives since the early 2000s: first, a triggering the various studiesa of the convergent and integrative dance; second, an attempt to the convergent and integrative program development research; third, the vitalization of the convergent research on dance digital contents; and fourth, a research on the convergent dance art phenomena. These researches have played an important role in boosting a change in the structure and realm expansion of dance studies that are interdisciplinary research enabling a holistic approach to the integration and convergence between scientific technique, skills of dance art and other studies. However, it should be acknowledged that one problem is the current research development plan or/and research program have very little feasibility and practicality except an interdisciplinary research on the dance digital contents. Therefore, it is suggested for the development of dance studies in the age of convergence as follows: first, a dance convergent study integrated in skills and theories of dance and science that would pave the way for an academic foundation leading to a new humanistic model in the age of convergent; and second, a need for a paradigm shift that theories should be deployed in the scene on a commercial scale in order to produce effectiveness of the interdisciplinary and integrative research on dance studies by turning into a behavioristic research phase. third, it needs to changeover from large scale of convergent performance into small scale of convergent performance based on original idea for accumulation of teachnique research and promotion of dance convergent performance.