• 한국전기전자재료학회논문지
• /
• 제19권9호
• /
• pp.800-807
• /
• 2006

To enhance the conversion speed more fast, we separate the determination process of MSB and LSB with the two independent ADC circuits of the Incremental Sigma Delta ADC. After the 1st Incremental Sigma Delta ADC conversion finished, the 2nd Incremental Sigma Delta ADC conversion start while the 1st Incremental Sigma Delta ADC work on the next input. By determining the MSB and the LSB independently, the ADC conversion speed is improved by two times better than the conventional Extended Counting Incremental Sigma Delta ADC. In processing the 2nd Incremental Sigma Delta ADC, the inverting sample/hold circuit inverts the input the 2nd Incremental Sigma Delta ADC, which is the output of switched capacitor integrator within the 1st Incremental Sigma Delta ADC block. The increased active area is relatively small by the added analog circuit, because the digital circuit area is more large than analog. In this paper, a 14 bit Extended Counting Incremental Sigma-Delta ADC is implemented in $0.25{\mu}m$ CMOS process with a single 2.5 V supply voltage. The conversion speed is about 150 Ksamples/sec at a clock rate of 25 MHz. The 1 MSB is 0.02 V. The active area is $0.50\;x\;0.35mm^{2}$. The averaged power consumption is 1.7 mW.

• ETRI Journal
• /
• 제34권4호
• /
• pp.645-648
• /
• 2012

An incremental subspace learning scheme to recover lost speech segments online is presented. Our contributions in this work are twofold. First, the recovery problem is transformed into an interpolation problem of the time-varying gains via nonnegative matrix factorization. Second, incremental nonnegative matrix factorization is employed to allow online processing and track the evolution of speech statistics. The effectiveness of the proposed scheme is confirmed by the experiment results.

• 정보처리학회논문지:컴퓨터 및 통신 시스템
• /
• 제3권3호
• /
• pp.73-80
• /
• 2014

본 논문에서는 이벤트구동 HDL 시뮬레이션에서 시뮬레이션 실행 시간 단축을 위한 인크리멘탈 시뮬레이션 방법을 제시한다. 일반적으로 시뮬레이션 과정은 일련의 반복적인 설계수정들과 동반되어 반복적으로 일어난다. 인크리멘탈 시뮬레이션은 이와같은 반복적인 시뮬레이션에서 설계수정 전의 시뮬레이션 결과를 이용하여서 설계수정 후에 진행되는 시뮬레이션의 수행 시간을 단축할 수 있는 효과적인 시뮬레이션 방법이다. 본 논문에서 제안된 인크리멘탈 시뮬레이션 방법의 유용함은 다양한 실제 디자인들에 적용한 실험을 통하여 확인할 수 있었다.

• 정보처리학회논문지:소프트웨어 및 데이터공학
• /
• 제9권3호
• /
• pp.77-82
• /
• 2020

차량에 탑재된 내비게이션의 대부분은 복잡한 구조의 PSF(Physical Storage Format) 파일 기반으로 개발되어 점증적 맵 갱신을 지원하기 어렵다. 이를 해결하기 위한 차세대 내비게이션 방법의 하나로서 DB 기반의 내비게이션 기술이 주목받고 있다. 점진적 맵 갱신을 지원하는 DB 기반 내비게이션 구현에 있어 지속적인 맵 데이터 갱신으로 인한 데이터 단편화현상으로 데이터 접근 비용이 증가할 수 있어 검색 성능의 저하가 발생할 수 있다. 본 논문에서는 점증적 맵 갱신을 지원하는 DB 기반 내비게이션의 성능 향상 방법의 하나로 데이터 단편화 방지 기법을 제시하고 실제 구현을 통하여 성능 향상 효과가 있음을 검증하였다.

• International Journal of Fuzzy Logic and Intelligent Systems
• /
• 제5권1호
• /
• pp.76-82
• /
• 2005

Speech signal has various features of speakers. This feature is extracted from speech signal processing. The speaker is identified by the speaker identification system. In this paper, we propose the speaker identification system that uses the incremental learning based on neural network. Recorded speech signal through the microphone is blocked to the frame of 1024 speech samples. Energy is divided speech signal to voiced signal and unvoiced signal. The extracted 12 orders LPC cpestrum coefficients are used with input data for neural network. The speakers are identified with the speaker identification system using the neural network. The neural network has the structure of MLP which consists of 12 input nodes, 8 hidden nodes, and 4 output nodes. The number of output node means the identified speakers. The first output node is excited to the first speaker. Incremental learning begins when the new speaker is identified. Incremental learning is the learning algorithm that already learned weights are remembered and only the new weights that are created as adding new speaker are trained. It is learning algorithm that overcomes the fault of neural network. The neural network repeats the learning when the new speaker is entered to it. The architecture of neural network is extended with the number of speakers. Therefore, this system can learn without the restricted number of speakers.

• 소성∙가공
• /
• 제17권1호
• /
• pp.27-34
• /
• 2008

In order to manufacture a doubly curved sheet metal, the incremental roll forming process which adopts advantages such as the flexibility of the incremental forming process and continuous bending deformation of the roll forming process has been experimentally investigated. An experimental equipment was developed which was named as unit roll set (URS) consisting of two pairs of support rolls and an upper center roll. The upper roll equipped with the servo control unit is motor-driven and can be positioned in the vertical direction according to the user's commands. Four support rolls are idle, and they freely rotate only along the axis so as to transfer the plate more stably in the tangential direction of the rotation of the driving roll. In the process, the plate is deformed incrementally as deformation proceeds simultaneously in longitudinal and transverse directions. Through the experiments using URS, information regarding to forming schedules is found out to fabricate curved hull plates. This study demonstrates the further application of the incremental roll forming process in shipbuilding industries.

• Journal of Information Processing Systems
• /
• 제5권2호
• /
• pp.105-116
• /
• 2009

To solve the general problems surrounding the application of genetic algorithms in stereo matching, two measures are proposed. Firstly, the strategy of simplified population-based incremental learning (PBIL) is adopted to reduce the problems with memory consumption and search inefficiency, and a scheme for controlling the distance of neighbors for disparity smoothness is inserted to obtain a wide-area consistency of disparities. In addition, an alternative version of the proposed algorithm, without the use of a probability vector, is also presented for simpler set-ups. Secondly, programmable graphics-hardware (GPU) consists of multiple multi-processors and has a powerful parallelism which can perform operations in parallel at low cost. Therefore, in order to decrease the running time further, a model of the proposed algorithm, which can be run on programmable graphics-hardware (GPU), is presented for the first time. The algorithms are implemented on the CPU as well as on the GPU and are evaluated by experiments. The experimental results show that the proposed algorithm offers better performance than traditional BMA methods with a deliberate relaxation and its modified version in terms of both running speed and stability. The comparison of computation times for the algorithm both on the GPU and the CPU shows that the former has more speed-up than the latter, the bigger the image size is.

• 정보처리학회논문지D
• /
• 제13D권6호
• /
• pp.797-806
• /
• 2006

데이터 웨어하우스에서는 질의를 빠르게 처리하기 위해 실체화 뷰(materialized view)가 흔히 사용된다. 실체화 뷰는 그의 정의에 포함된 데이터 소스들이 변경되면 이를 반영하기 위해 갱신되어야 한다. 실체화 년의 갱신은 많은 부하를 야기하므로, 실체화 뷰를 효율적으로 갱신하는 것은 매우 중요한 문제이다. 실체화 뷰의 효율적인 갱신 방법에 대해서는 이미 많은 연구가 있어왔지만, SPJ(Select-Project-Join) 형태로 정의된 실체화 뷰를 효율적으로 갱신하는 방법은 충분히 연구되지 않았다. 본 논문에서는 데이터 소스들에 대한 접근 비용을 최소화함으로써 SPJ 실체화 뷰를 효율적으로 점진적으로 갱신하는 방법을 제안한다. 제안하는 방법은 동적 계획법 알고리즘을 사용하여 최적의 갱신 방법을 찾는다. 마지막으로, 다양한 성능 평가 실험을 통해 제안하는 방법이 우수한 성능을 가지고 있음을 보인다.

• 정보처리학회논문지D
• /
• 제11D권3호
• /
• pp.533-544
• /
• 2004

항공관제 시스템, 워게임 등과 같이 시간제약을 갖는 대규모 트랜잭션 환경에서 보고서의 의미는 특별하다 이는 대규모 트랜잭션 연산을 수행하면서 성능의 제약 없이 제한된 시간 내에 보고서를 생성할 수 있어야 하기 때문이다. 이와 같이 대규모 트랜잭션 환경에서 시간제약을 만족하면서 보고서를 생성할 수 있도록 하기 위하여 이 논문에서는 점진적 연산 기법과 형성뷰 기법을 트리거와 저장 프로시져를 이용하여 결합시킨 모델을 제안한다. 아울러 제안 모델에 대한 구현 및 평가를 통해 제안 모델의 특성을 분석한다.

• International Journal of Contents
• /
• 제15권3호
• /
• pp.32-38
• /
• 2019

We usually suffer from difficulties in treating or managing Big Data generated from various digital media and/or sensors using traditional mining techniques. Additionally, there are many problems relative to the lack of memory and the burden of the learning curve, etc. in an increasing capacity of large volumes of text when new data are continuously accumulated because we ineffectively analyze total data including data previously analyzed and collected. In this paper, we propose a general-purpose classifier and its structure to solve these problems. We depart from the current feature-reduction methods and introduce a new scheme that only adopts changed elements when new features are partially accumulated in this free-style learning environment. The incremental learning module built from a gradually progressive formation learns only changed parts of data without any re-processing of current accumulations while traditional methods re-learn total data for every adding or changing of data. Additionally, users can freely merge new data with previous data throughout the resource management procedure whenever re-learning is needed. At the end of this paper, we confirm a good performance of this method in data processing based on the Big Data environment throughout an analysis because of its learning efficiency. Also, comparing this algorithm with those of NB and SVM, we can achieve an accuracy of approximately 95% in all three models. We expect that our method will be a viable substitute for high performance and accuracy relative to large computing systems for Big Data analysis using a PC cluster environment.