• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• fall
• /
• pp.21-22
• /
• 2021

최근 AI(Artificial Intelligence)를 기반으로 정밀한 가스 성분 감지를 위한 후각지능(Olfactory intelligence) 기술에 연구가 활발히 진행 중이다. 후각지능 학습데이터는 다른 감지 방식의 가스 센서들이 동시에 적용되는 멀티모달리티의 특성을 지니며 또한, 공간상에 분포된 센서 배열을 통해 획득된 다차원의 시계열 특성을 지닌다. 따라서 대량의 다차원 데이터에 대한 정확한 이해와 분석을 위해서는 데이터를 전처리하고 시각화할 수 있는 기술이 필요하다. 본 논문에서는 후각지능 학습을 위한 다차원의 복잡한 가스 데이터의 시각화를 위해 잡음 등의 불필요한 값을 제거하고, 데이터가 일관성을 가지도록 하며, 데이터의 차원을 시각화 가능하도록 축소하기 위한 전처리 방법을 제시한다.

• KNOM Review
• /
• v.24 no.1
• /
• pp.20-28
• /
• 2021

Extensive use of social media applications and mobile devices continues to increase data traffic. Social media applications generate an endless and massive amount of multimedia traffic, specifically video traffic. Many social media platforms such as YouTube, Daily Motion, and Netflix generate endless video traffic. On these platforms, only a few popular videos are requested many times as compared to other videos. These popular videos should be cached in the user vicinity to meet continuous user demands. MEC has emerged as an essential paradigm for handling consistent user demand and caching videos in user proximity. The problem is to understand how user demand pattern varies with time. This paper analyzes three publicly available datasets, MovieLens 20M, MovieLens 100K, and The Movies Dataset, to find the user request pattern over time. We find hourly, daily, monthly, and yearly trends of all the datasets. Our resulted pattern could be used in other research while generating and analyzing the user request pattern in MEC-based video caching scenarios.

• Proceedings of the Korean Society of Computer Information Conference
• /
• 2020.07a
• /
• pp.147-149
• /
• 2020

코로나바이러스의 세계 확산, 발병 이후 사람들의 실내활동 증가와 건강, 면역에 대한 관심은 많이 증가했다. 이에 맞춰 더욱 정교하고 바른 정보에 의한 스마트헬스케어 역시 관심이 증대되고 있다. 여기서 이야기하는 스마트헬스케어의 범위는 영상 장치를 비롯해 다양한 센서를 활용해 신체활동을 모니터링하고 분석하며 기존의 방식보다 더 객관적인 정보를 제공해 주는 것을 말한다. 위 기술과 대중의 관심을 바탕으로 하여 본 연구에서는 다중 모드 센서를 신체에 부착하여 신체활동을 모니터링 하는 시스템 개발을 목적으로 한다. 하드웨어 설계 부분에서 설계가 완성된 Arduino nano 33 Sense를 이용하여 스마트 헬스 실험 시간을 대폭 줄였다. 또한 운동과 같은 시계열 데이터를 분석하기 좋은 LSTM 기법을 채택하였으며, 개발된 모델을 추후 활용할 방안에 대해 논하였다.

• The Journal of the Korea Contents Association
• /
• v.5 no.4
• /
• pp.180-187
• /
• 2005

In this paper, we developed an online lecture for digital logic circuit which is a basic course in electric/electronic education. Because of importance of the laboratory experiences in this course and to reflect industrial requests, we have selected most effective experimental examples in each chapter and inserted instructions for basic usags of ORCAD and digial clock design. Moreover, we developed cyber lab to design students' own circuit using Flash animation. Two features of this cyber lab are real-like graphics for devices and breadboards to improve reality and patented new IC chip objects for easy experiments, which help the students understand digital logic easily.

• Journal of Korea Multimedia Society
• /
• v.8 no.1
• /
• pp.46-55
• /
• 2005

In this paper, we realize the panoramic infrared warning system to detect the small threaten object and propose the infrared image enhancement method to improve the warning ability of this system. This system composes of the sense head unit, the signal processing unit, and so on. In the proposed system, the sense head unit acquires the panoramic IR image with 360 degree field of view(FOV) by rotating the thermal sensor. The signal processing unit divides panoramic image into four sub-images with 90 degree FOV and computes the adaptive plateau value by using statistical characteristics of each subimage. Then the histogram equalization is performed for each subimage by using the adaptive plateau value. We realize the signal Processing unit by using the DSP and FPGA to perform the proposed method in real time. Experimental results show that the proposed method has better discrimination and lower false alarm rate than the conventional methods in this warning system.

• Journal of Korea Multimedia Society
• /
• v.4 no.6
• /
• pp.518-523
• /
• 2001

This paper proposes the new method which is adopted in extracting character features and recognizing characters using fractal dimension of the Chaos theory which highly recolonizes a minute difference with strange attractor created from Henon system. This paper implements a high precision character recognition system. firstly, it gets features of mesh, projection and cross distance feature from character images. And their feature is converted into data of time series. Then using modified Henon system suggested in this paper, each characters attractor about standard Korean Character, KSC 5601 is reconstructed. Secondly, in order to analyze the Chaotic degree of each characters attractor, it gets last features of character image after calculating box-counting Dimension, Natural Measure, Information Bit, Information Dimension which are meant fractal dimension. An experimental result shows 97.49% character classification rates for 2350 Korean characters using proposed method in this paper.

• The KIPS Transactions:PartD
• /
• v.15D no.4
• /
• pp.485-494
• /
• 2008

Huge amounts of various web items such as keywords, images, and web pages are being made widely available on the Web. The popularities of such web items continuously change over time, and mining temporal patterns in popularities of web items is an important problem that is useful for several web applications. For example, the temporal patterns in popularities of search keywords help web search enterprises predict future popular keywords, enabling them to make price decisions when marketing search keywords to advertisers. However, presence of millions of web items makes it difficult to scale up previous techniques for this problem. This paper proposes an efficient method for mining temporal patterns in popularities of web items. We treat the popularities of web items as time-series, and propose gapmeasure to quantify the similarity between the popularities of two web items. To reduce the computation overhead for this measure, an efficient method using the Fast Fourier Transform (FFT) is presented. We assume that the popularities of web items are not necessarily following any probabilistic distribution or periodic. For finding clusters of web items with similar popularity trends, we propose to use a density-based clustering algorithm based on the gap measure. Our experiments using the popularity trends of search keywords obtained from the Google Trends web site illustrate the scalability and usefulness of the proposed approach in real-world applications.