국제학술회의 참관기 - '95 국제신경회로망학술대회

채수익,Chae, Su-Ik 한국과학기술단체총연합회 1996 과학과 기술 Vol.29 No.2

본지는 국내외에서 활약하는 과학기술인들의 '국제학술회의 참관기'를 매월 연재한다. 이번호에는 호주 Perth에서 열린 '95 국제신경회로망학술대회'에 참석하고 돌아온 서울대 채수익교수의 참관기를 싣는다.

Hot Issue-MPSoC '04 Workshop을 다녀와서

채수익,Chae, Su-Ik IT-SOC협회 2004 IT SOC magazine Vol.2 No.-

프로젝트관리분야의 지능형 챗봇 서비스 사용의도에 영향을 미치는 요인 연구

채수익,최중영,조은영,이남용 한국IT정책경영학회 2021 한국IT정책경영학회 논문지 Vol.13 No.4

Recently, the IT technology of the 4th industrial revolution era, such as artificial intelligence technology and deep learning technology, is being activated, and the intelligent chatbot service based on it is being applied to the non-face-to-face work environment and its utilization is increasing. In this study, using the Technology Acceptance Model (TAM), we try to identify the factors that affect the intention to use the intelligent chatbot service in the project management field and verify the relationship between each factor. Data was collected through a survey, and verification was carried out using SPSS23 and AMOS23.0. Through this study, we intend to help introduce chatbot services to the project management field.

신속한 호 설정을 위한 위한 확장된 SIP 구조 제안

채수익(Soo-Ik Chae),김성용(Sung-Yong Kim),최현(Hyun Choi),박동선(Dong-Sun Park) 한국멀티미디어학회 2001 한국멀티미디어학회 학술발표논문집 Vol.2001 No.2

펄스열에서 1인 펄스수와 0인 펄스수의 비를 이용하여 확률연산을 하는 신경회로망

민승재,채수익 대한전자공학회 1994 전자공학회논문지-B Vol.b31 No.7

Stochastic computation employs random pulse streams to represent numbers. In this paper, we study a new method to implement the number system which uses the ratio of the numbers of ones and zeros in the pulse streams. In this number system. if P is the probability that a pulse is one in a pulse stream then the number X represented by the pulse stream is defined as P/(1-P). We propose circuits to implement the basic operations such as addition multiplication and sigmoid function with this number system and examine the error characteristics of such operations in stochastic computation. We also propose a neuron model and derive a learning algorithm based on backpropagation for the 3-layered feedforward neural networks. We apply this learning algorithm to a digit recognition problem. To analyze the results, we discuss the errors due to the variance of the random pulse streams and the quantization noise of finite length register.

로봇을 위한 운영체제 수준의 동적 전력 관리

최승민,채수익,Choi Seungmin,Chae Sooik 대한전자공학회 2005 電子工學會論文誌-SD (Semiconductor and devices) Vol.42 No.5

This paper describes a new approach for the operating system level power management to reduce the energy consumed in the IO devices in a robot platform, which provides various functions such as navigation, multimedia application, and wireless communication. The policy proposed in the paper, which was named the Energy-Aware Job Schedule (EAJS), rearranges the jobs scattered so that the idle periods of the devices are clustered into a time period and the devices are shut down during their idle period. The EAJS selects a schedule that consumes the minimum energyamong the schedules that satisfy the buffer and time constraints. Note that the burst job execution needs a larger memory buffer and causes a longer time delay from generating the job request until to finishing it. A prototype of the EAJS is implemented on the Linux kernel that manages the robot system. The experiment results show that a maximum $44\%$ power saving on a DSP and a wireless LAN card can be obtained with the EAJS. 본 논문에서는 가정용 로봇에 적용 할 수 있는 운영체제 수준의 동적 전력 관리 기법인 EAJS(Energy-Aware Job Scheduler)를 제안한다. EAJS는 workload가 일정하지 않은 IO 장치를 사용하는 job들을 스케줄하여 장치의 idle주기를 가능하면 연속적이게 만든 후, 장치를 shutdown 시켜서 에너지 소모를 줄이는 능동적인 저전력 스케줄러이다. EAJS는 기존의 저전력 스케줄러와 달리, IO 장치를 사용하는 job의 workload, job의 buffering에 사용 가능한 메모리의 크기, buffering으로 인해 발생하는 시간 지연 등을 동시에 고려하여 job을 스케줄 하기 때문에, 에너지를 절약하면서도 시스템의 성능 저하를 최소화 할 수 있다. EAJS의 prototype을 본 연구에서 개발한 가정용 로봇인 AFM(Autonomous Family Machine)에 구현하였으며, H.263 인코더를 수행시키는 실험을 통해 무선랜과 DSP의 에너지 소모가 최대 $44\%$ 가량 줄어듬을 확인 하였다.

초기 일반 지식을 갖고 있는 점증 적응 학습 알고리즘

오규환,채수익 대한전자공학회 1996 전자공학회논문지-B Vol.b33 No.2

This paper introduces the concept of fixed weights and proposes an algorithm for classification by adding this concept to vector space separation method in LVQ. The proposed algorithm is based on competitive learning. It uses fixed weightsfor generality and fast adaptation efficient radius for new weight creation, and L1 distance for fast calcualtion. It can be applied to many fields requiring adaptive learning with the support of generality, real-tiem processing and sufficient training effect using smaller data set. Recognition rate of over 98% for the train set and 94% for the test set was obtained by applying the suggested algorithm to on-line handwritten recognition.

국소적인 연결을 갖는 신경회로망을 이용한 스테레오 정합

박성진,채수익 대한전자공학회 1994 전자공학회논문지-B Vol.b31 No.12

In this paper, we propose a new approach of solving the stereopsis problem with a discrete-time cellular neural network(DTCNN) where each node has connections only with its local neithbors. Because the matching process of stereo correspondence depends on its geometrically local characteristics, the DTCNN is suitable for the stereo correspondence. Moreover, it can be easily implemented in VLSI. Therefore, we employed a two-layer DTCNN with dual templates, which are determined with the back propagation learning rule. Based on evaluation of the proposed approach for several random-dot stereograms, its performance is better than that of the Marr-Poggio algorithm.

부분 버스 반전 부호화를 이용한 시스템 수준 전력 최적화

신영수,채수익,최기영 대한전자공학회 1998 電子工學會論文誌, C Vol.c35 No.12

본 논문에서는 시스템 수준에서의 전력 최적화를 위한 한가지 방법으로 부분 버스 반전 부호화를 제안한다. 제안하는 방법에서는 버스 부호화가 버스선의 일부분에만 행해지는데, 이것은 버스선 전체를 부호화 함으로써 필요 없는 버스반전이 생기는 것을 피하기 위해서이다. 부분 버스 반전화를 위해서는 어떠한 버스선들을 선택하여 부호화하는가가 중요한데, 본 논문에서는 이러한 선택을 위한 알고리즘을 제안한다. 벤치마크 예제를 통한 모의 실험에서 부분 버스 반전 부호화 방법을 사용해 버스의 천이수를 평균 62.6% 줄일 수 있다는 것을 보였다. 또한 제안한 알고리즘의 성능을 평가하기 위해 시뮬레이티드 어닐링 방법과 비교하였다. We present a partial bus-invert coding scheme for system-level power optimization. In the proposed scheme, we select a sub-group of bus lines involved in bus encoding to avoid unnecessary inversion of bus lines not in the sub-group thereby reducing the total number of bus transitions. We propose a heuristic algorithm that selects the sub-group of bus lines for bus encoding. Experiments on benchmark examples indicate that the partial bus-invert coding reduces the total bus transitions by 62.6% on the average, compared to that of the unencoded patterns. We also compare the performance of the proposed heuristic algorithm with that of simulated annealing, which shows that it is highly efficient.

단층 신경망과 이중 기각 방법을 이용한 문자인식

임준호,채수익 대한전자공학회 1995 전자공학회논문지-B Vol.b32 No.3

Multilayer neural networks with backpropagation learning algorithm are widely used for pattern classification problems. For many real applications, it is more important to reduce the misclassification rate than to increase the rate of successful classification. But multilayer perceptrons(MLP's) have drawbacks of slow learning speed and false convergence to local minima. In this paper, we propose a new method for character recognition problems with a single-layer network and double rejection mechanisms, which guarantees a very low misclassification rate. Comparing to the MLP's, it yields fast learning and requires a simple hardware architecture. We also introduce a new coding scheme to reduce the misclassification rate. We have prepared two databases: one with 135,000 digit patterns and the other with 117,000 letter patterns, and have applied the proposed method for printed character recognition, which shows that the method reduces the misclassification rate significantly without sacrificing the correct recognition rate.