• Annual Conference on Human and Language Technology
• /
• 2017.10a
• /
• pp.35-40
• /
• 2017

기계 독해(Machine reading comprehension)는 주어진 문맥을 이해하고, 질문에 적합한 답을 문맥 내에서 찾는 문제이다. Simple Recurrent Unit (SRU)은 Gated Recurrent Unit (GRU)등과 같이 neural gate를 이용하여 Recurrent Neural Network (RNN)에서 발생하는 vanishing gradient problem을 해결하고, gate 입력에서 이전 hidden state를 제거하여 GRU보다 속도를 향상시킨 모델이며, Self-matching Network는 R-Net 모델에서 사용된 것으로, 자기 자신의 RNN sequence에 대하여 어텐션 가중치 (attention weight)를 계산하여 비슷한 의미 문맥 정보를 볼 수 있기 때문에 상호참조해결과 유사한 효과를 볼 수 있다. 본 논문에서는 한국어 기계 독해 데이터 셋을 구축하고, 여러 층의 SRU를 이용한 Encoder에 Self-matching layer를 추가한 $S^2$-Net 모델을 제안한다. 실험 결과, 본 논문에서 제안한 $S^2$-Net 모델이 한국어 기계 독해 데이터 셋에서 EM 65.84%, F1 78.98%의 성능을 보였다.

• 한국어정보학회:학술대회논문집
• /
• 2017.10a
• /
• pp.35-40
• /
• 2017

기계 독해(Machine reading comprehension)는 주어진 문맥을 이해하고, 질문에 적합한 답을 문맥 내에서 찾는 문제이다. Simple Recurrent Unit (SRU)은 Gated Recurrent Unit (GRU)등과 같이 neural gate를 이용하여 Recurrent Neural Network (RNN)에서 발생하는 vanishing gradient problem을 해결하고, gate 입력에서 이전 hidden state를 제거하여 GRU보다 속도를 향상시킨 모델이며, Self-matching Network는 R-Net 모델에서 사용된 것으로, 자기 자신의 RNN sequence에 대하여 어텐션 가중치 (attention weight)를 계산하여 비슷한 의미 문맥 정보를 볼 수 있기 때문에 상호참조해결과 유사한 효과를 볼 수 있다. 본 논문에서는 한국어 기계 독해 데이터 셋을 구축하고, 여러 층의 SRU를 이용한 Encoder에 Self-matching layer를 추가한 $S^2$-Net 모델을 제안한다. 실험 결과, 본 논문에서 제안한 $S^2$-Net 모델이 한국어 기계 독해 데이터 셋에서 EM 65.84%, F1 78.98%의 성능을 보였다.

• Proceedings of the KIEE Conference
• /
• 1996.07c
• /
• pp.1582-1584
• /
• 1996

The polycrystalline silicon thin film transistors (poly-Si TFT) are the most promising candidate for active matrix liquid crystal displays (AMLCD) for their high mobilities and current driving capabilities. The leakage current of the poly-Si TFT is much higher than that of the amorphous-Si TFT, thus larger storage capacitance is required which reduces the aperture ratio fur the pixel. The offset gated poly-Si TFTs have been widely investigated in order to reduce the leakage current. The conventional method for fabricating an offset device may require additional mask and photolithography process step, which is inapplicable for self-aligned source/drain ion implantation and rather cost inefficient. Due to mis-alignment, offset devices show asymmetric transfer characteristics as the source and drain are switched. We have proposed and fabricated a new offset poly-Si TFT by applying photoresist reflow process. The new method does not require an additional mask step and self-aligned ion implantation is applied, thus precise offset length can be defined and source/drain symmetric transfer characteristics are achieved.

• The Journal of the Acoustical Society of Korea
• /
• v.39 no.5
• /
• pp.414-423
• /
• 2020

In this paper, we propose a Dilated Convolution Gate Linear Unit (DCGLU) to mitigate the lack of sparsity and small receptive field problems caused by the segmentation map extraction process in sound event detection with weak labels. In the advent of deep learning framework, segmentation map extraction approaches have shown improved performance in noisy environments. However, these methods are forced to maintain the size of the feature map to extract the segmentation map as the model would be constructed without a pooling operation. As a result, the performance of these methods is deteriorated with a lack of sparsity and a small receptive field. To mitigate these problems, we utilize GLU to control the flow of information and Dilated Convolutional Neural Networks (DCNNs) to increase the receptive field without additional learning parameters. For the performance evaluation, we employ a URBAN-SED and self-organized bird sound dataset. The relevant experiments show that our proposed DCGLU model outperforms over other baselines. In particular, our method is shown to exhibit robustness against nature sound noises with three Signal to Noise Ratio (SNR) levels (20 dB, 10 dB and 0 dB).

• The Journal of Society for e-Business Studies
• /
• v.27 no.1
• /
• pp.15-28
• /
• 2022

Due to recent expansion of online market such as clothing, utilizing customer review has become a major marketing measure. User review has been used as a tool of analyzing sentiment of customers. Sentiment analysis can be largely classified with machine learning-based and lexicon-based method. Machine learning-based method is a learning classification model referring review and labels. As research of sentiment analysis has been developed, multi-modal models learned by images and video data in reviews has been studied. Characteristics of words in reviews are differentiated depending on products' and customers' categories. In this paper, sentiment is analyzed via considering review data and metadata of products and users. Gated Recurrent Unit (GRU), Long Short-Term Memory (LSTM), Self Attention-based Multi-head Attention models and Bidirectional Encoder Representation from Transformer (BERT) are used in this study. Same Multi-Layer Perceptron (MLP) model is used upon every products information. This paper suggests a multi-modal sentiment analysis model that simultaneously considers user reviews and product meta-information.

• The Journal of Korean Society for Radiation Therapy
• /
• v.22 no.2
• /
• pp.113-122
• /
• 2010

Purpose: To evaluate the accuracy of a target position at static and dynamic state by using Dynamic phantom for the difference between tumor's actual movement during respiratory gated radiation therapy and skin movement measured by RPM (Real-time Position Management). Materials and Methods: It self-produced Dynamic phantom that moves two-dimensionally to measure a tumor moved by breath. After putting marker block on dynamic phantom, it analyzed the amplitude and status change depending on respiratory time setup in advance by using RPM. It places marker block on dynamic phantom based on this result, inserts Gafchromic EBT film into the target, and investigates 5 Gy respectively at static and dynamic state. And it scanned investigated Gafchromic EBT film and analyzed dose distribution by using automatic calculation. Results: As a result of an analysis of Gafchromic EBT film's radiation amount at static and dynamic state, it could be known that dose distribution involving 90% is distributed within margin of error of 3 mm. Conclusion: As a result of an analysis of dose distribution's change depending on patient's respiratory cycle during respiratory gated radiation therapy, it is expected that the treatment would be possible within recommended margin of error at ICRP 60.

• Journal of the Korean Vacuum Society
• /
• v.5 no.2
• /
• pp.113-118
• /
• 1996

Aligning a cathode tip at the center of a gate hole is important in gated filed emission devices. We have fabricated a silicon field emitter using a following process so that a cathode and a gate hole are automatically aligned . After forming silicon tips on a silicon wafer, the wafer was covered with the $SiO_2$, gate metal, and photoresistive(PR) films. Because of the viscosity of the PR films, a spot where cathode tips were located protruded above the surface. By ashing the surface of the PR film, the gate metal above the tip apex was exposed when other area was still covered with the PR film. The exposed gate metal and subsequenlty the $SiO_2$ layer were selectively etched. The result produced a field emitter in which the gate film was in volcano shape and the cathode tip was located at the center of the gate hole. Computer simulation showed that the volcano shape and the cathode tip was located at the center of the gat hole. Computer simulation showed that the volcano shape emitter higher current and the electron beam which was focused better than the emitter for which the gate film was flat.

• International Journal of Advanced Culture Technology
• /
• v.12 no.3
• /
• pp.148-158
• /
• 2024

We investigate Autism Spectrum Disorder (ASD), which is typified by deficits in social interaction, repetitive behaviors, limited vocabulary, and cognitive delays. Traditional diagnostic methodologies, reliant on expert evaluations, frequently result in deferred detection and intervention, particularly in South Korea, where there is a dearth of qualified professionals and limited public awareness. In this study, we employ advanced deep learning algorithms to enhance early ASD screening through automated video analysis. Utilizing architectures such as Convolutional Long Short-Term Memory (ConvLSTM), Long-term Recurrent Convolutional Network (LRCN), and Convolutional Neural Networks with Gated Recurrent Units (CNN+GRU), we analyze video data from platforms like YouTube and TikTok to identify stereotypic behaviors (arm flapping, head banging, spinning). Our results indicate that the LRCN model exhibited superior performance with 79.61% accuracy on the augmented platform video dataset and 79.37% on the original SSBD dataset. The ConvLSTM and CNN+GRU models also achieved higher accuracy than the original SSBD dataset. Through this research, we underscore AI's potential in early ASD detection by automating the identification of stereotypic behaviors, thereby enabling timely intervention. We also emphasize the significance of utilizing expanded datasets from social media platform videos in augmenting model accuracy and robustness, thus paving the way for more accessible diagnostic methods.

• Journal of Internet Computing and Services
• /
• v.24 no.1
• /
• pp.39-47
• /
• 2023

The development of autonomous driving and Advanced Driver Assistance System (ADAS) technology has grown rapidly in recent years. As most traffic accidents occur due to human error, self-driving vehicles can drastically reduce the number of accidents and crashes that occur on the roads today. Obviously, technical advancements in autonomous driving can lead to improved public driving safety. However, due to the current limitations in technology and lack of public trust in self-driving cars (and drones), the actual use of Autonomous Vehicles (AVs) is still significantly low. According to prior studies, people's acceptance of an AV is mainly determined by trust. It is proven that people still feel much more comfortable in personalized ADAS, designed with the way people drive. Based on such needs, a new attempt for a customized ADAS considering each driver's driving style is proposed in this paper. Each driver's behavior is divided into two categories: assertive and defensive. In this paper, a novel customized ADAS algorithm with high classification accuracy is designed, which divides each driver based on their driving style. Each driver's driving data is collected and simulated using CARLA, which is an open-source autonomous driving simulator. In addition, Long Short-Term Memory (LSTM) and Gated Recurrent Unit (GRU) machine learning algorithms are used to optimize the ADAS parameters. The proposed scheme results in a high classification accuracy of time series driving data. Furthermore, among the vast amount of CARLA-based feature data extracted from the drivers, distinguishable driving features are collected selectively using Support Vector Machine (SVM) technology by comparing the amount of influence on the classification of the two categories. Therefore, by extracting distinguishable features and eliminating outliers using SVM, the classification accuracy is significantly improved. Based on this classification, the ADAS sensors can be made more sensitive for the case of assertive drivers, enabling more advanced driving safety support. The proposed technology of this paper is especially important because currently, the state-of-the-art level of autonomous driving is at level 3 (based on the SAE International driving automation standards), which requires advanced functions that can assist drivers using ADAS technology.

• Journal of the Korean Vacuum Society
• /
• v.7 no.2
• /
• pp.88-93
• /
• 1998

The fabrication process and emission characteristics of gated silicon field emitter arrays(FEAs) using chemical-mechanical-polishing (CMP) method are described. Novel fabrication techniques consisting of two-step dry etching with oxidation of silicon and CMP processes were developed for the formation of sharp tips and clear-cut edged gate electrodes, respectively. The gate height and aperture could be easily controlled by varying the polishing time and pressure in the CMP process. We obtained silicon FEAs having self-aligned and clear-cut edged gate electrode opening by eliminating the dishing problem during the CMP process with an oxide mask layer. The tip height of the finally fabricated FEAs was about 1.1 $\mu$m and the end radius of the tips was smaller than 100 $\AA$. The emission current meaured from the fabricated 2809 tips array was about 31 $\mu$A at a gate voltage of 80 V.