• 대한전자공학회:학술대회논문집
• /
• 대한전자공학회 1999년도 하계종합학술대회 논문집
• /
• pp.496-499
• /
• 1999

This paper describes a medular neural network(MNN) for a vision system which tracks a given object using a sequence of images from a camera unit. The MNN is used to precisely recognize the given robot endeffector and to minize the processing time. Since the robot endeffector can be viewed in many different shapes in 3-D space, a MNN structure, which contains a set of feedforwared neural networks, co be more attractive in recognizing the given object. Each single neural network learns the endeffector with a cluster of training patterns. The training patterns for a neural network share the similar charateristics so that they can be easily trained. The trained MNN is less sensitive to noise and it shows the better performance in recognizing the endeffector. The recognition rate of MNN is enhanced by 14% over the single neural network. A vision system with the MNN can precisely recognize the endeffector and place it at the center of a display for a remote operator.

• 정보처리학회논문지B
• /
• 제8B권6호
• /
• pp.699-704
• /
• 2001

본 논문에서는 혼성 모듈 구조의 recurrent 시간지연신경회로망(time-delay neural network)과 HMM(hidden Markov model)을 결합한 음성인식을 위한 새로운 구조에 대해 연구하였다. 시간지연신경회로망에서는 윈도우 크기를 확장하는 것이 인식률 향상에 유리하므로 이를 위해 첫 번째 은닉층에 궤환 구조를 사용하여 윈도우 크기를 실제로 크게 하지 않고도 동일한 효과를 얻을 수 있도록 하였다. 다음 이 시간지연신경망에서 입력된 음소의 특징 벡터의 시간에 따라 변화하는 성질을 잘 처리 할 수 있도록 시간지연신경회로망의 입력층을 복수의 상태로 나누어 음소특징의 시간축에 대한 각 상태마다 특징 감지기를 갖도록 하였다. 이때 시간지연신경회로망은 전체 음성인식 영역에 적용될 수 있도록 모듈 방식의 구조로 구성되었다. 그리고 이 모듈 구조 시간지연신경망의 출력 벡터를 HMM에 연결하여 서로 결합 하므로써 양 구조의 장점을 취하는 혼성 구조의 인식시스템을 구성하였고 이때 이 혼성 구조에서 효율적으로 적용할 수 있는 HMM 파라미터 smoothing 방법을 제시하였다.

• 전자공학회논문지A
• /
• 제30A권9호
• /
• pp.71-79
• /
• 1993

This paper proposes a new test pattern generation methodology for combinational logic circuits using neural networks based on a modular structure. The CUT (Circuit Under Test) is described in our gate level hardware description language. By conferring neural database, the CUT is compiled to an ATPG (Automatic Test Pattern Generation) neural network. Each logic gate in CUT is represented as a discrete Hopfield network. Such a neual network is called a gate module in this paper. All the gate modules for a CUT form an ATPG neural network by connecting each module through message passing paths by which the states of modules are transferred to their adjacent modules. A fault is injected by setting the activation values of some neurons at given values and by invalidating connections between some gate modules. A test pattern for an injected fault is obtained when all gate modules in the ATPG neural network are stabilized through evolution and mutual interactions. The proposed methodology is efficient for test generation, known to be NP-complete, through its massive paralelism. Some results on combinational logic circuits confirm the feasibility of the proposed methodology.

• 한국전기전자재료학회:학술대회논문집
• /
• 한국전기전자재료학회 2002년도 춘계학술대회 논문집 센서 박막재료 반도체재료 기술교육
• /
• pp.119-123
• /
• 2002

A new modular Lateral Information Propagation Networks(LIPN) has been designed. The LIPN has shown to be useful for reconstruction of information[3]. The problem is the fact that only the small number of nodes can be implemented in a IC chip with the circuit VLSI technology. The proposed modular architecture is propagated the neural network through inter module connections. For such inter module connections, the host (computer or logic) mediates the exchange of information among modules. Also border nodes in each module have capacitors for temporarily retaining the information from outer modules. The LIPN with $4{\times}4$ modules has been designed and simulation of interpolation with the designed LIPN has been done.

• 한국지능시스템학회:학술대회논문집
• /
• 한국퍼지및지능시스템학회 2003년도 ISIS 2003
• /
• pp.423-426
• /
• 2003

In this paper, we use the modular neural network and recurrent neural network structure to implement the artificial brain information processing. We also select related adaptive learning methods to learn the entirely new input in the existed neural network. With this, a part of information process in brain is implemented as and autonomous and adaptive model by neural network and further more, the entire model for information process in brain can be introduced.

• 전자공학회논문지CI
• /
• 제40권6호
• /
• pp.56-68
• /
• 2003

범용 신경망 연산기를 디지털 회로로 구현함에 있어 가장 까다로운 문제들 중 하나는 시냅스의 확장과 해당 네트워크에 맞게 뉴런들을 재배치하는 재구성 문제일 것이다. 본 논문에서는 이러한 문제들을 해결하기 위한 새로운 하드웨어 구조를 제안한다. 제안된 구조는 시냅스의 확장과 네트워크 구조의 변경을 위해 오리지날 디자인의 변경이 필요치 않으며, 모듈러 프로세싱 유니트의 확장을 통한 뉴런의 개수 및 레이어의 확장이 가능하다. 이 구조의 범용성 및 확장성에 대한 검증을 위해 다양한 종류의 다층 퍼셉트론 및 코호넨 네트워크를 구성하여 HDL 시뮬레이터를 통한 결과와 C 언어로 작성된 소프트웨어 시뮬레이터 결과를 비교하였으며 그 결과 성능이 거의 일치함을 확인하였다.

• 한국정보처리학회논문지
• /
• 제7권6호
• /
• pp.1834-1843
• /
• 2000

In this paper, we propose a modular neural network based classification method for handwritten numerals utilizing rotated and translated images of an input image. The whole numeral pattern space is divided into smaller spaces which overlap each other and form multiple clusters. On these multiple clusters, multiple multilayer perceptrons (MLP) neural networks, specialized in those clusters, are constructed. Thus, each MLP acts as an expert network on the corresponding cluster. An MLP is also used as a gating network functioning as a mediator among the multiple MLPs. In the learning phase, an input numeral image is dithered by tow geometric operations of translation and rotation so that new numeral images similar to original one are generated. In the recognition phase, we utilize not only input numeral image, but also nearly generated images through the rotation and the translation of the original image. Thus, multiple output values for those generated images were combined to make class decision by various combination methods. The experimental results confirm the validity of the proposed method.

• 제어로봇시스템학회:학술대회논문집
• /
• 제어로봇시스템학회 2001년도 ICCAS
• /
• pp.178.3-178
• /
• 2001

This paper is concerned with designing a neural network based navigator that is optimized in a user-defined sense for a mobile robot using ultrasonic sensors to travel to a goal position safely and efficiently without any prior map of the environment. The neural network has a dynamically reconfigurable structure that not only can optimize the weights but also the input sensory connectivity in order to meet any user-defined objective. Therefore, in this research, we can select an optimal subset of sensory inputs that results in the best performance related to both navigation and structural complexity. Further, this research uses the manually trained initial population and the modular neural network to alleviate ...

• 한국전기전자재료학회:학술대회논문집
• /
• 한국전기전자재료학회 2004년도 추계학술대회 논문집
• /
• pp.7-12
• /
• 2004

New modular Lateral Information Propagation Networks(LIPN) has been designed. The LIPN has shown to be useful for interpolation of information[3]. The problem is the fact that only the small number of nodes can be implemented in a IC chip with the circuit VLSI technology. The proposed modular architecture is for enlarging the neural network through inter module connections. For such inter module connections, the host(computer or logic) mediates the exchange of information among modules. Also border nodes in each module have capacitors for temporarily retaining the information from outer modules. Simulation of interpolation with the designed LIPN has been done through various experiments.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 1998년도 하계학술대회 논문집 G
• /
• pp.2206-2208
• /
• 1998

The modular Lateral Information Propagation Networks(LIPN) has been designed. The LIPN has shown to be useful for interpolation of information[3]. The problem is the fact that only the small number of nodes can be implemented in a IC chip with the circuit VLSI technology. The proposed modular architecture is for enlarging the neural network through inter module connections. For such inter module connections, the host(computer or logic) mediates the exchange of information among modules. Also border nodes in each module have capacitors for temporarily retaining the information from outer modules. The LIPN with $4{\times}4$ modules has been designed and simulation of interpolation with the designed LIPN has been done.