• Title/Summary/Keyword: Deep-Learning

Search Result 5,580, Processing Time 0.038 seconds

CAB: Classifying Arrhythmias based on Imbalanced Sensor Data

  • Wang, Yilin;Sun, Le;Subramani, Sudha
    • KSII Transactions on Internet and Information Systems (TIIS)
    • /
    • v.15 no.7
    • /
    • pp.2304-2320
    • /
    • 2021
  • Intelligently detecting anomalies in health sensor data streams (e.g., Electrocardiogram, ECG) can improve the development of E-health industry. The physiological signals of patients are collected through sensors. Timely diagnosis and treatment save medical resources, promote physical health, and reduce complications. However, it is difficult to automatically classify the ECG data, as the features of ECGs are difficult to extract. And the volume of labeled ECG data is limited, which affects the classification performance. In this paper, we propose a Generative Adversarial Network (GAN)-based deep learning framework (called CAB) for heart arrhythmia classification. CAB focuses on improving the detection accuracy based on a small number of labeled samples. It is trained based on the class-imbalance ECG data. Augmenting ECG data by a GAN model eliminates the impact of data scarcity. After data augmentation, CAB classifies the ECG data by using a Bidirectional Long Short Term Memory Recurrent Neural Network (Bi-LSTM). Experiment results show a better performance of CAB compared with state-of-the-art methods. The overall classification accuracy of CAB is 99.71%. The F1-scores of classifying Normal beats (N), Supraventricular ectopic beats (S), Ventricular ectopic beats (V), Fusion beats (F) and Unclassifiable beats (Q) heartbeats are 99.86%, 97.66%, 99.05%, 98.57% and 99.88%, respectively. Unclassifiable beats (Q) heartbeats are 99.86%, 97.66%, 99.05%, 98.57% and 99.88%, respectively.

Design and Implementation of Side-Type Finger Vein Recognizer (측면형 지정맥 인식기 설계 및 구현)

  • Kim, Kyeong-Rae;Choi, Hong-Rak;Kim, Kyung-Seok
    • The Journal of the Institute of Internet, Broadcasting and Communication
    • /
    • v.21 no.3
    • /
    • pp.159-168
    • /
    • 2021
  • As the information age enters, the use of biometrics using the body is gradually increasing because it is very important to accurately recognize and authenticate each individual's identity for information protection. Among them, finger vein authentication technology is receiving a lot of attention because it is difficult to forge and demodulate, so it has high security, high precision, and easy user acceptance. However, the accuracy may be degraded depending on the algorithm for identification or the surrounding light environment. In this paper, we designed and manufactured a side-type finger vein recognizer that is highly versatile among finger vein measuring devices, and authenticated using the deep learning model of DenseNet-201 for high accuracy and recognition rate. The performance of finger vein authentication technology according to the influence of the infrared light source used and the surrounding visible light was analyzed through simulation. The simulations used data from MMCBNU_6000 of Jeonbuk National University and finger vein images taken directly were used, and the performance were compared and analyzed using the EER.

Demand Prediction of Furniture Component Order Using Deep Learning Techniques (딥러닝 기법을 활용한 가구 부자재 주문 수요예측)

  • Kim, Jae-Sung;Yang, Yeo-Jin;Oh, Min-Ji;Lee, Sung-Woong;Kwon, Sun-dong;Cho, Wan-Sup
    • The Journal of Bigdata
    • /
    • v.5 no.2
    • /
    • pp.111-120
    • /
    • 2020
  • Despite the recent economic contraction caused by the Corona 19 incident, interest in the residential environment is growing as more people live at home due to the increase in telecommuting, thereby increasing demand for remodeling. In addition, the government's real estate policy is also expected to have a visible impact on the sales of the interior and furniture industries as it shifts from regulatory policy to the expansion of housing supply. Accurate demand forecasting is a problem directly related to inventory management, and a good demand forecast can reduce logistics and inventory costs due to overproduction by eliminating the need to have unnecessary inventory. However, it is a difficult problem to predict accurate demand because external factors such as constantly changing economic trends, market trends, and social issues must be taken into account. In this study, LSTM model and 1D-CNN model were compared and analyzed by artificial intelligence-based time series analysis method to produce reliable results for manufacturers producing furniture components.

Design of Real-time MR Contents using Substitute Videos of Vehicles and Background based on Black Box Video (블랙박스 영상 기반 차량 및 배경 대체 영상을 이용한 실시간 MR 콘텐츠의 설계)

  • Kim, Sung-Ho
    • Journal of Convergence for Information Technology
    • /
    • v.11 no.6
    • /
    • pp.213-218
    • /
    • 2021
  • In this paper, we detect and track vehicles by type based on highway daytime driving videos taken with black boxes for vehicles. In addition, we design a real-time MR contents production method that can be newly created by placing substitute videos of each type of detected vehicles in the same location as the new background video. To detect and track vehicles by type, we use the YOLO algorithm. And we also use the mask technique based on RGB color for substitute videos of each type of vehicles detected. The size of the vehicle substitute videos to be used for MR content are substituted by the same size as the area size of the detected vehicles. In this paper, we confirm that real-time MR contents design is possible as a result of experiments and simulations and believe that It will be usefully utilized in the field of VR contents.

Predicting win-loss using game data and deriving the importance of subdivided variables (게임데이터를 이용한 승패예측 및 세분화된 변수 중요도 도출 기법)

  • Oh, Min-Ji;Choi, Eun-Seon;Oui, Som Akhamixay;Cho, Wan-Sup
    • The Journal of Bigdata
    • /
    • v.5 no.2
    • /
    • pp.231-240
    • /
    • 2020
  • With the development in the IT industry and the growth in the game industry, user's game data is recorded in seconds according to various plays and options, and a vast amount of game data can be analyzed based on Bigdata. Combined with business, Bigdata is used to discover new values for profit creation in various fields, but it is utilized in the game industry in insufficient ways. In this study, considering the characteristics of the subdivided lines, we constructed a win-loss prediction model for each line using the game data of League of Legends, and derived the importance of variables. This study can contribute to planning of strategies for general game users to get information about team members in advance and increase the win rate by using the record search sites.

Recent Automatic Post Editing Research (최신 기계번역 사후 교정 연구)

  • Moon, Hyeonseok;Park, Chanjun;Eo, Sugyeong;Seo, Jaehyung;Lim, Heuiseok
    • Journal of Digital Convergence
    • /
    • v.19 no.7
    • /
    • pp.199-208
    • /
    • 2021
  • Automatic Post Editing(APE) is the study that automatically correcting errors included in the machine translated sentences. The goal of APE task is to generate error correcting models that improve translation quality, regardless of the translation system. For training these models, source sentence, machine translation, and post edit, which is manually edited by human translator, are utilized. Especially in the recent APE research, multilingual pretrained language models are being adopted, prior to the training by APE data. This study deals with multilingual pretrained language models adopted to the latest APE researches, and the specific application method for each APE study. Furthermore, based on the current research trend, we propose future research directions utilizing translation model or mBART model.

Implementation of a Classification System for Dog Behaviors using YOLI-based Object Detection and a Node.js Server (YOLO 기반 개체 검출과 Node.js 서버를 이용한 반려견 행동 분류 시스템 구현)

  • Jo, Yong-Hwa;Lee, Hyuek-Jae;Kim, Young-Hun
    • Journal of the Institute of Convergence Signal Processing
    • /
    • v.21 no.1
    • /
    • pp.29-37
    • /
    • 2020
  • This paper implements a method of extracting an object about a dog through real-time image analysis and classifying dog behaviors from the extracted images. The Darknet YOLO was used to detect dog objects, and the Teachable Machine provided by Google was used to classify behavior patterns from the extracted images. The trained Teachable Machine is saved in Google Drive and can be used by ml5.js implemented on a node.js server. By implementing an interactive web server using a socket.io module on the node.js server, the classified results are transmitted to the user's smart phone or PC in real time so that it can be checked anytime, anywhere.

Development of personalized clothing recommendation service based on artificial intelligence (인공지능 기반 개인 맞춤형 의류 추천 서비스 개발)

  • Kim, Hyoung Suk;Lee, Jong Hyuck;Lee, Hyun Dong
    • Smart Media Journal
    • /
    • v.10 no.1
    • /
    • pp.116-123
    • /
    • 2021
  • Due to the rapid growth of the online fashion market and the resulting expansion of online choices, there is a problem that the seller cannot directly respond to a large number of consumers individually, although consumers are increasingly demanding for more personalized recommendation services. Images are being tagged as a way to meet consumer's personalization needs, but when people tagging, tagging is very subjective for each person, and artificial intelligence tagging has very limited words and does not meet the needs of users. To solve this problem, we designed an algorithm that recognizes the shape, attribute, and emotional information of the product included in the image with AI, and codes this information to represent all the information that the image has with a combination of codes. Through this algorithm, it became possible by acquiring a variety of information possessed by the image in real time, such as the sensibility of the fashion image and the TPO information expressed by the fashion image, which was not possible until now. Based on this information, it is possible to go beyond the stage of analyzing the tastes of consumers and make hyper-personalized clothing recommendations that combine the tastes of consumers with information about trends and TPOs.

An Analysis of Existing Studies on Parallel and Distributed Processing of the Rete Algorithm (Rete 알고리즘의 병렬 및 분산 처리에 관한 기존 연구 분석)

  • Kim, Jaehoon
    • The Journal of Korean Institute of Information Technology
    • /
    • v.17 no.7
    • /
    • pp.31-45
    • /
    • 2019
  • The core technologies for intelligent services today are deep learning, that is neural networks, and parallel and distributed processing technologies such as GPU parallel computing and big data. However, for intelligent services and knowledge sharing services through globally shared ontologies in the future, there is a technology that is better than the neural networks for representing and reasoning knowledge. It is a knowledge representation of IF-THEN in RIF or SWRL, which is the standard rule language of the Semantic Web, and can be inferred efficiently using the rete algorithm. However, when the number of rules processed by the rete algorithm running on a single computer is 100,000, its performance becomes very poor with several tens of minutes, and there is an obvious limitation. Therefore, in this paper, we analyze the past and current studies on parallel and distributed processing of rete algorithm, and examine what aspects should be considered to implement an efficient rete algorithm.

Fault Diagnosis of Bearing Based on Convolutional Neural Network Using Multi-Domain Features

  • Shao, Xiaorui;Wang, Lijiang;Kim, Chang Soo;Ra, Ilkyeun
    • KSII Transactions on Internet and Information Systems (TIIS)
    • /
    • v.15 no.5
    • /
    • pp.1610-1629
    • /
    • 2021
  • Failures frequently occurred in manufacturing machines due to complex and changeable manufacturing environments, increasing the downtime and maintenance costs. This manuscript develops a novel deep learning-based method named Multi-Domain Convolutional Neural Network (MDCNN) to deal with this challenging task with vibration signals. The proposed MDCNN consists of time-domain, frequency-domain, and statistical-domain feature channels. The Time-domain channel is to model the hidden patterns of signals in the time domain. The frequency-domain channel uses Discrete Wavelet Transformation (DWT) to obtain the rich feature representations of signals in the frequency domain. The statistic-domain channel contains six statistical variables, which is to reflect the signals' macro statistical-domain features, respectively. Firstly, in the proposed MDCNN, time-domain and frequency-domain channels are processed by CNN individually with various filters. Secondly, the CNN extracted features from time, and frequency domains are merged as time-frequency features. Lastly, time-frequency domain features are fused with six statistical variables as the comprehensive features for identifying the fault. Thereby, the proposed method could make full use of those three domain-features for fault diagnosis while keeping high distinguishability due to CNN's utilization. The authors designed massive experiments with 10-folder cross-validation technology to validate the proposed method's effectiveness on the CWRU bearing data set. The experimental results are calculated by ten-time averaged accuracy. They have confirmed that the proposed MDCNN could intelligently, accurately, and timely detect the fault under the complex manufacturing environments, whose accuracy is nearly 100%.