• ETRI Journal
• /
• v.36 no.2
• /
• pp.253-263
• /
• 2014

In the near future, traditional narrow and fixed viewpoint video services will be replaced by high-quality panorama video services. This paper proposes a visual-attention-aware progressive region of interest (RoI) trick mode streaming service (VA-PRTS) that prioritizes video data to transmit according to the visual attention and transmits prioritized video data progressively. VA-PRTS enables the receiver to speed up the time to display without degrading the perceptual quality. For the proposed VA-PRTS, this paper defines a cutoff visual attention metric algorithm to determine the quality of the encoded video slice based on the capability of visual attention and the progressive streaming method based on the priority of RoI video data. Compared to conventional methods, VA-PRTS increases the bitrate saving by over 57% and decreases the interactive delay by over 66%, while maintaining a level of perceptual video quality. The experiment results show that the proposed VA-PRTS improves the quality of the viewer experience for interactive panoramic video streaming services. The development results show that the VA-PRTS has highly practical real-field feasibility.

• Annual Conference on Human and Language Technology
• /
• 2020.10a
• /
• pp.9-13
• /
• 2020

관계 추출은 주어진 문장이나 문서에 존재하는 개체들 간의 의미적 관계를 찾아내는 작업을 말한다. 최근 문서 수준 관계 추출 말뭉치인 DocRED가 공개되면서 문서 수준 관계 추출에 대한 연구가 활발히 진행되고 있다. 또한 사전 학습된 Masked Language Model(MLM)이 자연어처리 분야 전체에 영향력을 보이면서 관계 추출에서도 MLM을 사용하는 연구가 진행되고 있다. 그러나 문서 수준의 관계 추출은 문서의 단위가 길기 때문에 Self-attention을 기반으로 하는 MLM을 사용하면 모델의 계산량이 증가하는 문제가 있다. 본 논문은 이 점을 보완하기 위해 관계 추출에 필요한 문장을 선별하는 간단한 전처리 방법을 제안한다. 또한 문서의 길이에 상관없이 관계 추출에 필요한 어휘 정보를 자동으로 습득 할 수 있는 Relation-Context Co-attention 방법을 제안한다. 제안 모델은 DocRED 말뭉치에서 Dev F1 62.01%, Test F1 59.90%로 높은 성능을 보였다.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2022.06a
• /
• pp.1311-1314
• /
• 2022

DNN을 사용하여 객체 인식 과정에서 객체를 잘 분류하기 위해서는 시각적 설명성이 요구된다. 시각적 설명성은 object class에 대한 예측을 pixel-wise attribution으로 표현해 예측 근거를 해석하기 위해 제안되었다, Scale-invariant한 특징을 제공하도록 설계된 pyramidal features 기반 backbone 구조는 object detection 및 classification 등에서 널리 쓰이고 있으며, 이러한 특징을 갖는 feature pyramid를 trainable attention mechanism에 적용하고자 할 때 계산량 및 메모리의 복잡도가 증가하는 문제가 있다. 본 논문에서는 일반적인 FPN에서 객체 인식 성능과 설명성을 높이기 위한 피라미드-주의집중 계층네트워크 (FPN-Attention Layered Network) 방식을 제안하고, 실험적으로 그 특성을 평가하고자 한다. 기존의 FPN만을 사용하였을 때 객체 인식 과정에서 설명성을 향상시키는 방식이 객체 인식에 미치는 정도를 정량적으로 평가하였다. 제안된 모델의 적용을 통해 낮은 computing 오버헤드 수준에서 multi-level feature를 고려한 시각적 설명성을 개선시켜, 결괴적으로 객체 인식 성능을 향상 시킬 수 있음을 실험적으로 확인할 수 있었다.

• Human Ecology Research
• /
• v.58 no.2
• /
• pp.267-278
• /
• 2020

This study examined the longitudinal effect of preschool children's media exposure and maternal depression on first-grade children's school adjustment and the mediating effect of attention problem. Longitudinal data from the Panel Study of Korean Children (PSKC) collected by the Korea Institute of Child Care and Education (KICCE) was used to examine this hypothetical model. The subjects of the study included 2,150 children (1,091 boys and 1,059 girls) and their mothers across 2013 (5 yrs.) through 2015 (7 yrs.). The Structural Equation Model (SEM) was estimated using SPSS 25.0 and Amos 25. The results of this study were as follows. First, higher level of preschool children's media exposure and maternal depression were related to higher attention problems after a year and lower level of children's school adjustment during first-grade. Second, preschool children's media exposure and maternal depression had an indirect effect on first-grade children's school adjustment via attention problem. The results of this study will provide supporting evidence to many educators and parents for the implementation of effective practices for first-grade children to enhance their school adjustment. Furthermore, this study emphasizes the importance of maternal psychological wellbeing and the risk of indiscriminate media exposure during early childhood on first-grade's school adjustment.

• International Journal of Contents
• /
• v.15 no.4
• /
• pp.8-15
• /
• 2019

This paper presents a solution of the 'Quick, Draw! Doodle Recognition Challenge' hosted by Google. Doodles are drawings comprised of concrete representational meaning or abstract lines creatively expressed by individuals. In this challenge, a doodle is presented as a sequence of sketches. From the view of at the sketch level, to learn the pattern of strokes representing a doodle, we propose a sequential model stacked with multiple convolution layers and Long Short-Term Memory (LSTM) cells following the attention mechanism [15]. From the view at the image level, we use multiple models pre-trained on ImageNet to recognize the doodle. Finally, an ensemble and a post-processing method using the minimum cost flow algorithm are introduced to combine multiple models in achieving better results. In this challenge, our solutions garnered 11th place among 1,316 teams. Our performance was 0.95037 MAP@3, only 0.4% lower than the winner. It demonstrates that our method is very competitive. The source code for this competition is published at: https://github.com/ngxbac/Kaggle-QuickDraw.

• Human Ecology Research
• /
• v.51 no.4
• /
• pp.371-382
• /
• 2013

The purpose of this study is to investigate children's emotional problems (attention problem, aggression, somatic symptom, social withdrawal, depression) and to examine the relationship between children's emotional problems with family characteristics (parent education, parent job, family income), life-styles (gaming times, TV times, sleep time). The sample was 2,140 collections of second grade children and their parents who participated in Korea Youth Panel Survey on 2011. We analyzed the data which were collected by means of questionnaires and the data were analyzed with t -test, ANOVA, Pearson correlation analysis, and regression analysis with SPSS ver. 19.0. The results were summarized as follows. The level of children's emotional problem was relatively low on average. There is a significant difference in the children's emotional problems according to family characteristics and life-styles. There was a relationship between children's emotional problems (attention problem, aggression, somatic symptom, social withdrawal, depression) and life-styles (gaming times, TV times). The significant factors influencing the children's emotional problems are connected with father's education, mother's job, family income, family characteristics, gaming time and TV time of life style. It is noted that parent education level was an important factor for children's attention problem and aggression. Amongst children's life-styles, gaming time and TV time are negative factors for social withdrawal and depression.

• The Journal of the Convergence on Culture Technology
• /
• v.10 no.1
• /
• pp.501-506
• /
• 2024

Humans have been shown to engage in multitasking behavior when searching for information on two or more topics or searching an information system at the same time. When processing multiple information tasks, priorities must be established as there are cognitive and physical limitations in processing multiple information tasks at once. The level of cognitive attention involved in multitasking behavior can vary depending on the complexity and importance of the information task. The objectives of this study are to understand: (a) the relationship between attention and information task prioritization behavior when people interact with information retrieval systems to find information for multiple tasks; (b) The effect of the degree of attention on information task prioritization behavior when people interact with an IR system to find information for multiple tasks. A review of the relevant literature shows that when people interact with information retrieval systems to find information for multiple tasks, their level of attention affects how they prioritize multiple information tasks. It should be noticed that people pay more attention to things they find interesting or important. Human-centered system design based on a conceptual understanding of multitasking is discussed.

• Journal of Korea Society of Industrial Information Systems
• /
• v.26 no.1
• /
• pp.1-10
• /
• 2021

This study proposed a Long-Short Term Memory network to consider changes in emotion over time, and applied an attention mechanism to give weights to the emotion states that appear at specific moments. We used 32 channel EEG data from DEAP database. A 2-level classification (Low and High) experiment and a 3-level classification experiment (Low, Middle, and High) were performed on Valence and Arousal emotion model. As a result, accuracy of the 2-level classification experiment was 90.1% for Valence and 88.1% for Arousal. The accuracy of 3-level classification was 83.5% for Valence and 82.5% for Arousal.

• Journal of East Asia Management
• /
• v.5 no.1
• /
• pp.1-31
• /
• 2024

Based on the perspective of consumers and the method of value engineering, this paper uses "CSR expectation deviate level" to measure corporate social responsibility, and discusses the influence of corporate social responsibility on financial performance and its action path. This paper collected the questionnaire survey data of 878 consumers and the panel data of 98 listed companies from 2009 to 2012. The empirical results show that: (1) Consumers pay more attention to products and services, charity, environmental protection and their responsibilities to employees, and less attention to their responsibilities to shareholders or creditors and partners; (2) Corporate social responsibility is negatively correlated with financial performance, and corporate marketing ability plays a moderating role in it. That is, the smaller the gap between the level of corporate social responsibility fulfilled by enterprises and consumers' expectations, the better the financial performance of enterprises, which also reminds enterprises that they need to rationally allocate corporate social responsibility resources and constantly cultivate their own marketing capabilities, so as to better meet the level of corporate social responsibility expected by consumers. The value engineering method quantifies consumers' value perception of corporate social responsibility, which has a certain practical guiding role. Of course, there are some limitations in this paper, and future research can further explore the potential impact mechanism.

• Journal of The Korean Society of Integrative Medicine
• /
• v.2 no.2
• /
• pp.13-20
• /
• 2014

Purpose : The purpose of this study was to evaluate the influence of cognition and executive function by computer - based attention program in vascular dementia. Method : The subjects of this study, old man diagnosed with vascular dementia, 12 patients were picked up, who were agreed with this research and were having hospital care for 4 weeks at nursing care centers. Computer-based attention program was applied to vascular dementia. Cognitive function measured by a K-MMSE and executive function measured by ACL. The SPSS Ver. 18.0 statistical program was used for data processing. The significance level for statistical inspection was set as 0.05. Result : In comparison of cognitive function was not significantly correlated in the pre and post test and executive function was significantly correlated in the pre and post test. Conclusion : Therefore, the computer-based attention program is useful to improve the cognitive and executive function in elderly with vascular dementia.