• Annual Conference on Human and Language Technology
• /
• 2017.10a
• /
• pp.125-129
• /
• 2017

채팅 시스템은 사람이 사용하는 자연어를 이용해 컴퓨터와 대화를 하는 시스템이다. 한국어 특성상 대화체에서 동일한 의미를 가졌지만 다른 형태를 가진 경우가 많다. 본 논문에서는 Attention mechanism Encoder-Decoder Model을 사용해 한국어 특성에 맞는 효과적인 생성 모델을 만들 수 있는 입력, 출력 단위를 제안한다. 실험에서 정성 평가와 ROUSE, BLEU 평가를 진행한 결과 형태소 단위의 입력 보다 본 논문에서 제안한 색인어 입력 단위의 성능이 높고, 의사 형태소 단위 출력 보다 음절 단위 출력을 사용한 시스템이 더 문법적 오류가 적고 적합한 응답을 생성하는 것을 보였다.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2016.06a
• /
• pp.264-266
• /
• 2016

During the $2^{nd}$ JVET (Joint Group on Future Video Coding Technology Exploration) meeting, up to 22 coding tools focusing on Future Video Coding (FVC) were proposed. Despite that the application of proposed coding tools has a considerable performance enhancement, however, the encoding time of Joint Exploration Model (JEM) software is over 20 times for All Intra coding mode, 6 times for Random Access coding mode, of HEVC reference model (HM), and decoding time is 1.6 times for All Intra coding mode, 7.9 times for Random Access coding mode, of HM. This paper focuses on analyzing the complexity of the JEM software compared with HM.

• Journal of Biomedical Engineering Research
• /
• v.24 no.1
• /
• pp.45-53
• /
• 2003

Based on the 4-compartmental pharmacokinetic model developed in PART1, target-controlled infusion(TCI) pump system was designed and evaluated. The TCI system consists of digital board including microcontroller and digital signal process(DSP), analog board, motor-driven actuator, user friendly interface, power management and controller. It provides two modes according to the drugs: plasma target concentration and effect target concentration. Anaesthetist controls the depth of anaesthesia for patients by adjusting the required concentration to maintain both plasma and effect site in drug concentration. The data estimated in DSP include infusion rate, initial load dose, and rotation number of motor encoder. During TCI operation, plasma concentration. effect site concentration, awaken concentration, context-sensitive decrement time and system error information are displayed in real time. Li-ion battery guarantees above 2 hours without power line failure. For high reliability of the system, two microprocessors were used to perform independent functions for both pharmacokinetic algorithm and motor control strategy.

• 한국어정보학회:학술대회논문집
• /
• 2017.10a
• /
• pp.125-129
• /
• 2017

채팅 시스템은 사람이 사용하는 자연어를 이용해 컴퓨터와 대화를 하는 시스템이다. 한국어 특성상 대화체에서 동일한 의미를 가졌지만 다른 형태를 가진 경우가 많다. 본 논문에서는 Attention mechanism Encoder-Decoder Model을 사용해 한국어 특성에 맞는 효과적인 생성 모델을 만들 수 있는 입력, 출력 단위를 제안한다. 실험에서 정성 평가와 ROUSE, BLEU 평가를 진행한 결과 형태소 단위의 입력 보다 본 논문에서 제안한 색인어 입력 단위의 성능이 높고, 의사 형태소 단위 출력 보다 음절 단위 출력을 사용한 시스템이 더 문법적 오류가 적고 적합한 응답을 생성하는 것을 보였다.

• 전기의세계
• /
• v.30 no.11
• /
• pp.717-727
• /
• 1981

The phase-locked loop technique developed in the 1930's has many advantages when applied to speed control. The speed control and analysis of a three phase induction motor using the PLL are described in this paper. In this system, the phase frequency detector (PFD) compares the actual motor speed from the pulses received from a shaft encoder and desired speed, and the difference adjusts the frequency of the inverter that feeds the motor, and excellent speed regulation in the order of 0.035(%) has been-obtained. A linear continuous model of the drive is developed and system response is analysed using conventional root locus techniques. Various compensating filters and feedback signals are considered and the need for addition of derivative feedback is shown. A sampled data model is used to study the effects of discrete PFD output. Stability limitson speed are predicted. A drive was implimented and experimental results are presented to verify theoretical predictions.

• Proceedings of the IEEK Conference
• /
• 2002.06e
• /
• pp.181-184
• /
• 2002

In this paper, an algorithm is proposed for the automatic movement of the platform on a ladder lift which operates by using hydraulic force. Experimental data of the manual movement operated by experts are acquired and analyzed. The length of movement of the platform is estimated by the number of pulses from an encoder while the platform is moving. The operating characteristics have been extracted from the data. The number of pulses varies whenever the weight of load varies. We propose a model to compensate the difference of the number of pulses. An automatic movement algorithm based on the model is proposed. The platform can start and stop smoothly at the starting point and the destination point. Experimental results show that the error distance at the destination point is less than 5 cm when the length of the ladder is 23 m.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.13 no.8
• /
• pp.4255-4269
• /
• 2019

This paper proposes an image generation method using a Multi Discriminator Generative Adversarial Net (MDGAN) as a next generation 2D game sprite creation technique. The proposed GAN is an Autoencoder-based model that receives three areas of information-color, shape, and animation, and combines them into new images. This model consists of two encoders that extract color and shape from each image, and a decoder that takes all the values of each encoder and generates an animated image. We also suggest an image processing technique during the learning process to remove the noise of the generated images. The resulting images show that 2D sprites in games can be generated by independently learning the three image attributes of shape, color, and animation. The proposed system can increase the productivity of massive 2D image modification work during the game development process. The experimental results demonstrate that our MDGAN can be used for 2D image sprite generation and modification work with little manual cost.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.13 no.1
• /
• pp.385-405
• /
• 2019

High Efficiency Video Coding (HEVC) enables significantly improved compression performance relative to existing standards. However, the advance also requires high computational complexity. To accelerate the intra prediction mode decision, a minimum risk Bayesian classification framework is introduced. The classifier selects a small number of candidate modes to be evaluated by a rate-distortion optimization process using the sum of absolute Hadamard transformed difference (SATD). Moreover, the proposed method provides a loss factor that is a good trade-off model between computational complexity and coding efficiency. Experimental results show that the proposed method achieves a 31.54% average reduction in the encoding run time with a negligible coding loss of 0.93% BD-rate relative to HEVC test model 16.6 for the Intra_Main common test condition.

• Journal of Advanced Information Technology and Convergence
• /
• v.10 no.1
• /
• pp.45-56
• /
• 2020

Despite extensive research, performance enhancement of keyphrase (KP) extraction remains a challenging problem in modern informatics. Recently, deep learning-based supervised approaches have exhibited state-of-the-art accuracies with respect to this problem, and several of the previously proposed methods utilize Bidirectional Encoder Representations from Transformers (BERT)-based language models. However, few studies have investigated the effective application of BERT-based fine-tuning techniques to the problem of KP extraction. In this paper, we consider the aforementioned problem in the context of scientific articles by investigating the fine-tuning characteristics of two distinct BERT models - BERT (i.e., base BERT model by Google) and SciBERT (i.e., a BERT model trained on scientific text). Three different datasets (WWW, KDD, and Inspec) comprising data obtained from the computer science domain are used to compare the results obtained by fine-tuning BERT and SciBERT in terms of KP extraction.

• Journal of the Korean Society of Industry Convergence
• /
• v.25 no.4_2
• /
• pp.707-714
• /
• 2022

Versatile Video Coding (VVC) is the most recent video coding standard, which had been developed by Joint Video Expert Team (JVET). It can improve significant coding performance compared to the previous standard, namely High Efficiency Video Coding (HEVC). Although VVC can achieve the powerful coding performance, it requires the tremendous computational complexity of VVC encoder. Especially, affine motion compensation (AMC) was adopted the block-based 4-parameter or 6-parameter affine prediction to overcome the limit of translational motion model while VVC require the cost of higher encoding complexity. In this paper, we proposed the early termination of AMC that determines whether the affine motion estimation for AMC is performed or not. Experimental results showed that the proposed method reduced the encoding complexity of affine motion estimation (AME) up to 16% compared to the VVC Test Model 17 (VTM17).