• Electronics and Telecommunications Trends
• /
• v.35 no.3
• /
• pp.20-33
• /
• 2020

Temporal action detection (TAD) in untrimmed videos is an important but a challenging problem in the field of computer vision and has gathered increasing interest recently. Although most studies on action in videos have addressed action recognition in trimmed videos, TAD methods are required to understand real-world untrimmed videos, including mostly background and some meaningful action instances belonging to multiple action classes. TAD is mainly composed of temporal action localization that generates temporal action proposals, such as single action and action recognition, which classifies action proposals into action classes. However, the task of generating temporal action proposals with accurate temporal boundaries is challenging in TAD. In this paper, we discuss TAD technologies that are considered high performance in terms of representative TAD studies based on deep learning. Further, we investigate evaluation methodologies for TAD, such as benchmark datasets and performance measures, and subsequently compare the performance of the discussed TAD models.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.15 no.8
• /
• pp.2993-3010
• /
• 2021

The job shop scheduling problem (JSSP) plays a critical role in smart manufacturing, an effective JSSP scheduler could save time cost and increase productivity. Conventional methods are very time-consumption and cannot deal with complicated JSSP instances as it uses one optimal algorithm to solve JSSP. This paper proposes an effective scheduler based on deep learning technology named self-supervised long-short term memory (SS-LSTM) to handle complex JSSP accurately. First, using the optimal method to generate sufficient training samples in small-scale JSSP. SS-LSTM is then applied to extract rich feature representations from generated training samples and decide the next action. In the proposed SS-LSTM, two channels are employed to reflect the full production statues. Specifically, the detailed-level channel records 18 detailed product information while the system-level channel reflects the type of whole system states identified by the k-means algorithm. Moreover, adopting a self-supervised mechanism with LSTM autoencoder to keep high feature extraction capacity simultaneously ensuring the reliable feature representative ability. The authors implemented, trained, and compared the proposed method with the other leading learning-based methods on some complicated JSSP instances. The experimental results have confirmed the effectiveness and priority of the proposed method for solving complex JSSP instances in terms of make-span.

• The Journal of the Korea Contents Association
• /
• v.6 no.5
• /
• pp.85-95
• /
• 2006

In this paper, we propose a method for simplifying the modeling of 'walk' for various instances of quadruped and easily extending the model of 'walk' for the instance of new subordinate class which is not defined in the original class hierarchy by using the sophisticatedly modeled 'walk' of quadruped. To achieve this method, we apply the analyzed pattern of walk and classification according to the structural similarity of quadruped studied in the previous researches to the class hierarchy and propose a method for inheriting the actions of super class. This paper model the 'walk' of quadruped by concertizing the characteristics of quadruped and defining the necessary factors and appropriate domains in terms of parameters of 'walk' and apply the model to the horse and cow, typical instances of quadruped.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.17 no.2
• /
• pp.7-13
• /
• 2017

Many systems rely on reliable timestamps to determine the time of a particular action or event. This is especially true in digital investigations where investigators are attempting to determine when a suspect actually committed an action. The challenge, however, is that objects are not updated at the exact moment that an event occurs, but within some time-span after the actual event. In this work we define a simple model of digital systems with objects that have associated timestamps. The model is used to predict object update patterns for objects with associated timestamps, and make predictions about these update time-spans. Through empirical studies of digital systems, we show that timestamp update patterns are not instantaneous. We then provide a method for calculating the distribution of timestamp updates on a particular system to determine more accurate action instance times.

• Proceedings of the Korea Contents Association Conference
• /
• 2004.11a
• /
• pp.564-569
• /
• 2004

Most existing ontologies are suited for static knowledge, so they lack capability of representing procedural knowledge which is essential for an agent's action control or inference in the virtual world. Also they are not prepared to describe objects beyond their fixed ranges of domains as designed. In this paper we propose cyber-microcosm ontology (CMO) which augments procedural aspects and expressive power for multiple forms rather than fixed form as in conventional ontologies. The resulting ontology will provide an enhanced knowledge structure to capture procedural aspects of agents' actions and to facilitate their associated inferencing. The procedural aspects of the CMO are designed based on action frame formed according to diverse elements. They are elaborated in terms of various qualifiers and quantifiers to reflect statistical natures over time and instances.

• Korean journal of applied entomology
• /
• v.22 no.2 s.55
• /
• pp.147-159
• /
• 1983

In general, herbicides have been classified according to selectivity, mobility. time of application, methods of application, mode of action and chemical property and structure. However, there was no generally accepted classification system for practical use in the field. The primary processes affected by the majority of herbicides are the growth process through cell elongation and/or cell division, the photosynthetic process specifically the light reaction, the oxidative phosphorylation and the integrity of the membrane systems. The usual approach in the study of the mechanism by which herbicides kill or inhibit the growth of plants is to initially determine the morphological phototoxicity systems, The mechanism by which a herbicide kills a plant or suppresses its development is actually the resultant effect of primary and secondary(or side) effects. In most instances, the death of the plant is due to the secondary effects. To induce the desired response, a herbicide must be able to gain entry into the plants and once inside, to be transported within the plant to its site(s) of activity in concentrations great enough. Obstacles to the entry and movement of herbicides in plants are generally classified by leaf and soil obstacles, translocation obstacles and biochemical obstacles, and these obstacles are also strongly influenced by plant species and by environmental factors such as light, temperature, rainfall and relative humidity. And hence, in most instances, results obtained from laboratory or greenhous vary from those of field experiment. Author attempted to classify herbicides from the field experiment using the two-dimensional ordination analysis to obtain practical information for selecting effective herbicides or to choose effective herbicide combinations for increasing herbicidal efficacy or reducing the chemical cost. Based on this two-dimensional diagram, desired herbicides or combinations were selected and further investigated for the interaction effects whether these combinations are synergistic, additive or antagonistic. From the results, it was concluded that these new approach could possibly be give more comprehensive informations about effective use of herbicide than any other systems.

• Journal of Internet Computing and Services
• /
• v.22 no.4
• /
• pp.13-28
• /
• 2021

Reinforcement Learning (RL) is a machine learning algorithm that repeat the closed-loop process that agents perform actions specified by the policy, the action is evaluated with a reward function, and the policy gets updated accordingly. The key benefit of RL is the ability to optimze the policy with action evaluation. Hence, it can effectively be applied to developing advanced intelligent systems and autonomous systems. Conventional RL incoporates a single policy, a reward function, and relatively simple policy update, and hence its utilization was limited. In this paper, we propose an extended RL model that considers multiple instances of RL elements. We define a formal model of the key elements and their computing model of the extended RL. Then, we propose design methods for applying to system development. As a case stud of applying the proposed formal model and the design methods, we present the design and implementation of an advanced car navigator system that guides multiple cars to reaching their destinations efficiently.

• BMB Reports
• /
• v.45 no.3
• /
• pp.177-182
• /
• 2012

Camellia japonica oil (CJ oil) has been used traditionally in East Asia to nourish and soothe the skin as well as help restore the elasticity of skin. CJ oil has also been used on all types of bleeding instances. However, little is known about its anti-inflammatory effects. Therefore, the anti-inflammatory effects of CJ oil and its mechanisms of action were investigated. CJ oil inhibited LPS-induced production of NO, $PGE_2$, and TNF-${\alpha}$ in RAW264.7 cells. In addition, expression of COX-2 and iNOS genes was reduced. To evaluate the mechanism of the anti-inflammatory activity of CJ oil, LPS-induced activation of AP-1 and NF-${\kappa}B$ promoters was found to be significantly reduced by CJ oil. LPS-induced phosphorylation of $I{\kappa}B{\alpha}$, ERK, p38, and JNK was also attenuated. Our results indicate that CJ oil exerts anti-inflammatory effects by downregulating the expression of iNOS and COX-2 genes through inhibition of NF-${\kappa}B$ and AP-1 signaling.

• International conference on construction engineering and project management
• /
• 2022.06a
• /
• pp.465-472
• /
• 2022

Manual material handling is the one of the leading causes for musculoskeletal disorders (MSDs) and lower back discomfort. According to a study, construction formworkers suffer greater rates of muscular injuries and related illness due to manual activities. However, there is still a paucity of information on MSD, preventive posture issues, and corresponding solutions for construction aluminum formworkers. As a result, MSD and disregard of worker health and safety continue to exist at construction sites. Although preventive measures and strategies have been studied in previous research, we believe it is imperative to shed light on this problem through this study. This study aims to 1) implement a simple and cost-effective elevated bench to reduce MSDs, and 2) determine the rapid upper limbs assessment (RULA) and Ovako working posture analyzing system (OWAS) action catagory of workers in different postures to assess their MSD conditions and obtain an optimal position and posture using the Jack human modeling software and simulation tool. The study findings reveal a considerable reduction in MSD discomfort and which posture is acceptable in post-intervention instances.Thus results provide inexpensive and simple ergonomic interventions with favorable RULA and OWAS ratings that can be applied at construction sites. This study demonstrates workstation ergonomic intervention cases that can aid in understanding the urgency of applying existing research strategies into practice.

• The Korean Journal of Pharmacology
• /
• v.31 no.2
• /
• pp.153-164
• /
• 1995

Sympathomimetic amines, especially ephedrine, are a major ingredient in proprietary medications for symptomatic treatment of upper respiratory infections. Their frequent uses can lead to occasional instances of abuse and habituation. The clinical symptoms of ephedrine abuse are similar to that of amphetamine psychosis and resemble closely that of schizophrenia. Because both amphetamine psychosis and schizophrenia are thought to be mediated primarily through the action on catecholamines, ephedrine-induced changes of the biogenic amines can be suspected. However, there were few studies about the central effects of ephedrine because of the milder central action than peripheral. Therefore, the present investigation was undertaken to elucidate the relations between the effects of single or repeated administration of ephedrine on the regional levels of biogenic amines in rat brain and ephedrine-induced CNS stimulation. The male Sprague-Dawley rats weighing $100{\sim}200\;g$ were used. After single or repeated administrations of ephedrine, blocks of tissue were obtained from frontal cortex, corpus striatum, hippocampus, thalamus, hypothalamus, substantia nigra and cerebellum. The concentration of biogenic amines(norepinephrine, epinephrine, dopamine, 5-hydroxytryptamine(5-HT)) and their metabolites (3,4-dihydroxyphenylacetic acid (DOPAC), homovanillic acid(HVA), 5-hydroxyindoleacetic acid(5-HIAA)) were measured by means of high performance liquid chromatography-electrochemical detector(HPLC-ECD). The results obtained were as follows: 1) In the normal rat, the concentration of norepinephrine was the highest in hypothalamus. Dopamine, DOPAC and HVA were highest in corpus striatum, and 5-HT and 5-HIAA were highest in substantia nigra. Epinephrine was not detectable in any part of the brain tissue. 2) In a single administration of ephedrine, the concentration of DOPAC was decreased in corpus striatum. However, the other biogenic amines and their metabolites were not changed. 3) In repeated administration of ephedrine, the concentration of norepinephrine was decreased in all brain region checked. Dopamine was decreased in corpus striatum and substantia nigra and, increased in hypothalamus, and HVA was decreased in corpus striatum. 5-HT was decreased in all brain region except cerebellum and, 5-HIAA was decreased only in frontal cortex. The ratio of 5-HIAA/5-HT was increased in corpus striatum, thalamus, hypothalamus and substantia nigra. These data indicated that, although a single administration of ephedrine did not change the central neurotransmitters, repeated administration of ephedrine caused the decreases of norepinephrine and 5-HT in the most regions of brain, which may be responsible for the emergence of abnormal behavioral effect after ephedrine abuse.