• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.21 no.3
• /
• pp.107-114
• /
• 2021

Recently artificial intelligence technology has been introduced in various fields and various machine learning models have been operated in various frameworks as academic interest has increased. However, these frameworks have different data formats, which lack interoperability, and to overcome this, the open neural network exchange format, ONNX, has been proposed. In this paper we describe how to transform multiple machine learning models to ONNX, and propose algorithms and inference systems that can determine machine learning techniques in an integrated ONNX format. Furthermore we compare the inference results of the models before and after the ONNX transformation, showing that there is no loss or performance degradation of the learning results between the ONNX transformation.

• KIPS Transactions on Software and Data Engineering
• /
• v.10 no.3
• /
• pp.109-114
• /
• 2021

GAN(Generative Adversarial Network) is an image generation model, which is composed of a generator network and a discriminator network, and generates an image similar to a real image. Since the image generated by the GAN should be similar to the actual image, a loss function is used to minimize the loss error of the generated image. However, there is a problem that the loss function of GAN degrades the quality of the image by making the learning to generate the image unstable. To solve this problem, this paper analyzes GAN-related studies and proposes an edge GAN(eGAN) using edge detection. As a result of the experiment, the eGAN model has improved performance over the existing GAN model.

• Annual Conference on Human and Language Technology
• /
• 2023.10a
• /
• pp.366-371
• /
• 2023

신경망 기반의 검색 모델이 활발히 연구됨에 따라 효과적인 대조학습을 위한 다양한 네거티브 샘플링 방법이 제안되고 있다. 대표적으로, ANN전략은 하드 네거티브 샘플링 방법으로 질문에 대해 검색된 후보 문서들 중에서 정답 문서를 제외한 상위 후보 문서를 네거티브로 사용하여 검색 모델의 성능을 효과적으로 개선시킨다. 하지만 질문에 부착된 정답 문서를 통해 후보 문서를 네거티브로 구분하기 때문에 실제로 정답을 유추할 수 있는 후보 문서임에도 불구하고 네거티브로 분류되어 대조학습을 진행할 수 있다는 문제점이 있다. 이러한 가짜 네거티브 문제(False Negative Problem)는 학습과정에서 검색 모델을 혼란스럽게 하며 성능을 감소시킨다. 본 논문에서는 False Negative Problem를 분석하고 이를 완화시키기 위해 가짜 네거티브 분류기(False Negative Classifier)를 소개한다. 실험은 오픈 도메인 질의 응답 데이터셋인 Natural Question에서 진행되었으며 실제 False Negative를 확인하고 이를 판별하여 기존 성능보다 더 높은 성능을 얻을 수 있음을 보여준다.

• Journal of KIISE:Software and Applications
• /
• v.34 no.12
• /
• pp.1065-1074
• /
• 2007

Among various methods to compare reference signatures with an input signature, the segment-to-segment matching method has more advantages than global and point-to-point methods. However, the segment-to-segment matching method has the problem of having lower recognition rate according to the variation of partitioning points. To resolve this drawback, this paper proposes a signature verification method by considering linear discriminant analysis as well as segment-to-segment matching method. For the final decision step, we adopt statistical based Bayesian classifier technique to effectively combine two individual systems. Under the various experiments, the proposed method shows better performance than segment-to-segment based matching method.

• The KIPS Transactions:PartB
• /
• v.11B no.7 s.96
• /
• pp.885-892
• /
• 2004

In this paper, we propose a multi-modal biometrics system based on face and signature recognition system. Here, the face recognition system is designed by fuzzy LDA, and the signature recognition system is implemented with the LDA and segment matching methods. To effectively aggregate two systems, we obtain statistical distribution models based on matching values for genuine and impostor, respectively. And then, the final verification is Performed by the support vector machine. From the various experiments, we find that the proposed method shows high recognition rates comparing with the conventional methods.

• Proceedings of the KSRS Conference
• /
• 2007.03a
• /
• pp.161-166
• /
• 2007

Multi-functional Transport Satellite lR(MTSAT-lR)과 같은 정지궤도 기상위성의 지상 전처리 과정에는 영상위치보정(Image navigation and registration)이 포함된다. 영상위치보정은 위성 영상의 기하학적인 왜곡을 보정하는 과정이다. 랜드마크를 이용하는 영상위치보정 과정은 랜드마크 결정과 센서 모델 추정， 리샘플링(Resampling)의 세 가지 단계로 나눌 수 있다. MTSAT-1R의 High Resolution Image Data(HiRID)는 이미 영상위치보정이 수행되었지만, 기하학적인 오차가 남아있는 영상을 포함하기도 한다. 본 연구에서는 이런 기하학적인 오차를 제거하기 위해서 강인추정 기법에 기반한 기하보정을 수행하였다. 이태윤 등 (2005)은 강인추정 기법과 Direct Linear Transformation (DLT)에 기반한 오정합 판별 방법을 제안하였다. 이 판별 방법을 적용하여 추정된 DLT로 MTSAT-1R 영상의 기하보정을 수행한 결과에는 향상된 정확도로 기하보정 된 영상 뿐만 아니라 비교적 큰 오차를 포함하는 영상도 있었다. 이를 해결하기 위해서 본 연구에서는 강인추정 기법과 Affine 변환을 이용한 방법을 적용하였다. 본 연구에서는 기준 해안선에서 추출한 1,407개의 랜드마크와 8개의 MTSAT-1R 영상을 이용하였으며，강인추정 기법에 DLT를 적용한 방법과 Affine 변환을 적용한 방법으로 자동 기하보정을 수행하여 그 결과를 비교하였다. 또한 강인추정 기볍 중 RANSAC과 MSAC의 적용 결과를 비교하여 보았다. 그 결과，DLT로 기하보정 시，본 논문에서 제안된 방법이 강인추정 기법에 DLT를 적용한 방법 보다 더 좋은 성능을 보여주었다.

• The Journal of Korean Association of Computer Education
• /
• v.13 no.6
• /
• pp.99-104
• /
• 2010

USN Fire Prevention System is an intelligent system that detects the fire through the value which has got from a sensor such as temperature, humidity, intensity of illumination, acceleration, carbon dioxide(CO2) and so on. And then send it to the operator also use the algorithmic fire detection to operate fire extinguish system on. It is among U-Disaster Prevention System which has prevented fire lately. Configuration of the packet was designed to make the most of lightweight and fast processing for low power consumption. Recently listed in the encryption algorithm is applied each DES, 3DES, AES and HIGHT. So objective was to faster encryption than encryption of high-performance finally domestic standard encryption algorithm HIGHT were suitable for the fire prevention system needed frequent sensing time.

• Journal of the Korean Electrochemical Society
• /
• v.23 no.4
• /
• pp.81-89
• /
• 2020

To understand the performance of the electrochemical device, the analysis of the mechanism of ionic conduction is important. However, due to the ionic interaction in the electrolyte and the complexity of the electrolyte structure, a clear analysis method of the ion conduction mechanism has not been proposed. Instead, a variety of mathematical models have been devised to explain the mechanism of ion conduction, and this review introduces the Arrhenius and Vogel-Tammann-Fulcher (VTF) model. In general, the above two mathematical models are used to describe the temperature dependence of the transport properties of electrolytes such as ionic conductivity, diffusion coefficient, and viscosity, and a suitable model can be determined through the linearity of the graph consisting of the logarithm of the moving property and the reciprocal of the temperature. Currently, many electrolyte studies are evaluating the suitability of the above two models for electrolytes by varying the composition and temperature range, and the ion conduction mechanism analysis and activation energy calculation are in progress. However, since there are no models that can accurately describe the transport properties of electrolytes, new models and improvement of existing models are needed.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.23 no.4
• /
• pp.41-46
• /
• 2023

Human urine is a process of excreting waste products in the blood, and it is easy to collect and contains various substances. Urinalysis is used to check for diseases, health conditions, and urinary tract infections. There are three methods of urinalysis: physical property test, chemical test, and microscopic test, and chemical test results can be easily confirmed using urine test strips. A variety of items can be tested on the urine test strip, through which various diseases can be identified. Recently, with the spread of smart phones, research on reading urine test strips using smart phones is being conducted. There is a method of detecting and reading the color change of a urine test strip using a smartphone. This method uses the RGB values and the color difference formula to discriminate. However, there is a problem in that accuracy is lowered due to various environmental factors. This paper applies a deep learning model to solve this problem. In particular, color discrimination of a urine test strip is improved in a smartphone using a lightweight CNN (Convolutional Neural Networks) model. CNN is a useful model for image recognition and pattern finding, and a lightweight version is also available. Through this, it is possible to operate a deep learning model on a smartphone and extract accurate urine test results. Urine test strips were taken in various environments to prepare deep learning model training images, and a urine test service application was designed using MobileNet V3.

• Journal of KIISE:Software and Applications
• /
• v.27 no.6
• /
• pp.678-689
• /
• 2000

This paper describes an automatic summarization approach that constructs a summary by extracting sentences that are likely to represent the main theme of a document. As a way of selecting summary sentences, the system uses a model that takes into account lexical and statistical information obtained from a document corpus. As such, the system consists of two parts: the training part and the summarization part. The former processes sentences that have been manually tagged for summary sentences and extracts necessary statistical information of various kinds, and the latter uses the information to calculate the likelihood that a given sentence is to be included in the summary. There are at least three unique aspects of this research. First of all, the system uses a text component identification model to categorize sentences into one of the text components. This allows us to eliminate parts of text that are not likely to contain summary sentences. Second, although our statistically-based model stems from an existing one developed for English texts, it applies the framework to individual features separately and computes the final score for each sentence by combining the pieces of evidence using the Dempster-Shafer combination rule. Third, not only were new features introduced but also all the features were tested for their effectiveness in the summarization framework.