• Journal of the Korea Society of Computer and Information
• /
• v.27 no.10
• /
• pp.1-9
• /
• 2022

In this paper, we propose the Deep Learning-Based Companion Animal Abnormal Behavior Detection Service, which using video and sensor data. Due to the recent increase in households with companion animals, the pet tech industry with artificial intelligence is growing in the existing food and medical-oriented companion animal market. In this study, companion animal behavior was classified and abnormal behavior was detected based on a deep learning model using various data for health management of companion animals through artificial intelligence. Video data and sensor data of companion animals are collected using CCTV and the manufactured pet wearable device, and used as input data for the model. Image data was processed by combining the YOLO(You Only Look Once) model and DeepLabCut for extracting joint coordinates to detect companion animal objects for behavior classification. Also, in order to process sensor data, GAT(Graph Attention Network), which can identify the correlation and characteristics of each sensor, was used.

• Journal of Internet of Things and Convergence
• /
• v.10 no.4
• /
• pp.9-19
• /
• 2024

In this paper, a real-time public transportation monitoring system is proposed. The proposed system was implemented by developing a public transportation app and utilizing optical sensors, pressure sensors, and an object detection algorithm. Additionally, a bus model was created to verify the system's functionality. The proposed real-time public transportation monitoring system has three key features. First, the app can monitor congestion levels within public transportation by detecting seat occupancy and the total number of passengers based on changes in optical and pressure sensor readings. Second, to prevent errors in the optical sensor that can occur when multiple passengers board or disembark simultaneously, we explored the possibility of using the YOLO object detection algorithm to verify the number of passengers through CCTV footage. Third, convenience is enhanced by displaying occupied seats in different colors on a separate screen. The system also allows users to check their current location, available public transportation options, and remaining time until arrival. Therefore, the proposed system is expected to offer greater convenience to public transportation users.

• Smart Structures and Systems
• /
• v.26 no.4
• /
• pp.495-506
• /
• 2020

Monitoring of complex processes faces several challenges mainly due to the lack of relevant sensory information or insufficient elaborated decision-making strategies. These challenges motivate researchers to adopt complex data processing and analysis in order to improve the process representation. This paper presents the development and implementation of quality monitoring framework based on a model-driven approach using embedded artificial intelligence strategies. In this work, the strategies are applied to the supervision of a microfabrication process aiming at showing the great performance of the framework in a very complex system in the manufacturing sector. The procedure involves two methods for modelling a representative quality variable, such as surface roughness. Firstly, the hybrid incremental modelling strategy is applied. Secondly, a generalized fuzzy clustering c-means method is developed. Finally, a comparative study of the behavior of the two models for predicting a quality indicator, represented by surface roughness of manufactured components, is presented for specific manufacturing process. The manufactured part used in this study is a critical structural aerospace component. In addition, the validation and testing are performed at laboratory and industrial levels, demonstrating proper real-time operation for non-linear processes with relatively fast dynamics. The results of this study are very promising in terms of computational efficiency and transfer of knowledge to manufacturing industry.

• Journal of Digital Convergence
• /
• v.19 no.9
• /
• pp.257-269
• /
• 2021

The purpose of this study is to collect biosignal data in a non-invasive and non-restrictive manner using a BCG (Ballistocardiogram) sensor, and utilize artificial intelligence machine learning algorithms in ICT and high-performance computing environments. And it is to present and study a method for developing and validating a data-based blood glucose prediction model. In the blood glucose level prediction model, the input nodes in the MLP architecture are data of heart rate, respiration rate, stroke volume, heart rate variability, SDNN, RMSSD, PNN50, age, and gender, and the hidden layer 7 were used. As a result of the experiment, the average MSE, MAE, and RMSE values of the learning data tested 5 times were 0.5226, 0.6328, and 0.7692, respectively, and the average values of the validation data were 0.5408, 0.6776, and 0.7968, respectively, and the coefficient of determination (R²) was 0.9997. If research to standardize a model for predicting blood sugar levels based on data and to verify data set collection and prediction accuracy continues, it is expected that it can be used for non-invasive blood sugar level management.

• The korean journal of orthodontics
• /
• v.51 no.2
• /
• pp.77-85
• /
• 2021

Objective: To evaluate the accuracy of a multi-stage convolutional neural network (CNN) model-based automated identification system for posteroanterior (PA) cephalometric landmarks. Methods: The multi-stage CNN model was implemented with a personal computer. A total of 430 PA-cephalograms synthesized from cone-beam computed tomography scans (CBCT-PA) were selected as samples. Twenty-three landmarks used for Tweemac analysis were manually identified on all CBCT-PA images by a single examiner. Intra-examiner reproducibility was confirmed by repeating the identification on 85 randomly selected images, which were subsequently set as test data, with a two-week interval before training. For initial learning stage of the multi-stage CNN model, the data from 345 of 430 CBCT-PA images were used, after which the multi-stage CNN model was tested with previous 85 images. The first manual identification on these 85 images was set as a truth ground. The mean radial error (MRE) and successful detection rate (SDR) were calculated to evaluate the errors in manual identification and artificial intelligence (AI) prediction. Results: The AI showed an average MRE of 2.23 ± 2.02 mm with an SDR of 60.88% for errors of 2 mm or lower. However, in a comparison of the repetitive task, the AI predicted landmarks at the same position, while the MRE for the repeated manual identification was 1.31 ± 0.94 mm. Conclusions: Automated identification for CBCT-synthesized PA cephalometric landmarks did not sufficiently achieve the clinically favorable error range of less than 2 mm. However, AI landmark identification on PA cephalograms showed better consistency than manual identification.

• ETRI Journal
• /
• v.44 no.3
• /
• pp.438-449
• /
• 2022

This study proposes a dependency grammar-based self-attention multilayered bidirectional long short-term memory (DG-M-Bi-LSTM) model for subject-predicate-object (SPO) tuple recognition from natural language (NL) sentences. To add recent knowledge to the knowledge base autonomously, it is essential to extract knowledge from numerous NL data. Therefore, this study proposes a high-accuracy SPO tuple recognition model that requires a small amount of learning data to extract knowledge from NL sentences. The accuracy of SPO tuple recognition using DG-M-Bi-LSTM is compared with that using NL-based self-attention multilayered bidirectional LSTM, DG-based bidirectional encoder representations from transformers (BERT), and NL-based BERT to evaluate its effectiveness. The DG-M-Bi-LSTM model achieves the best results in terms of recognition accuracy for extracting SPO tuples from NL sentences even if it has fewer deep neural network (DNN) parameters than BERT. In particular, its accuracy is better than that of BERT when the learning data are limited. Additionally, its pretrained DNN parameters can be applied to other domains because it learns the structural relations in NL sentences.

• Computers and Concrete
• /
• v.34 no.1
• /
• pp.93-122
• /
• 2024

One well-known reason for using Fiber Reinforced Polymer (FRP) composites is to improve concrete strength and strain capacity via external confinement. Hence, various studies have been undertaken to offer a good illustration of the response of FRP-wrapped concrete for practical design intents. However, in such studies, the strength and strain of the confined concrete were predicted using regression analysis based on a limited number of test data. This study presents an approach based on artificial neural networks (ANNs) to develop models to predict the strength and strain at maximum stress enhancement of circular concrete cross-sections confined with different FRP types (Carbone, Glass, Aramid). To achieve this goal, a large test database comprising 493 axial compression experiments on FRP-confined concrete samples was compiled based on an extensive review of the published literature and used to validate the predicted artificial intelligence techniques. The ANN approach is currently thought to be the preferred learning technique because of its strong prediction effectiveness, interpretability, adaptability, and generalization. The accuracy of the developed ANN model for predicting the behavior of FRP-confined concrete is commensurate with the experimental database compiled from published literature. Statistical measures values, which indicate a better fit, were observed in all of the ANN models. Therefore, compared to existing models, it should be highlighted that the newly developed models based on FRP type are remarkably accurate.

• Journal of Sensor Science and Technology
• /
• v.29 no.1
• /
• pp.19-26
• /
• 2020

Electrocardiogram (ECG) classification has become an essential task of modern day wearable devices, and can be used to detect cardiovascular diseases. State-of-the-art Artificial Intelligence (AI)-based ECG classifiers have been designed using various artificial neural networks (ANNs). Despite their high accuracy, ANNs require significant computational resources and power. Herein, three different ANNs have been compared: multilayer perceptron (MLP), convolutional neural network (CNN), and spiking neural network (SNN) only for the ECG classification. The ANN model has been developed in Python and Theano, trained on a central processing unit (CPU) platform, and deployed on a PYNQ-Z2 FPGA board to validate the model using a Jupyter notebook. Meanwhile, the hardware accelerator is designed with Overlay, which is a hardware library on PYNQ. For classification, the MIT-BIH dataset obtained from the Physionet library is used. The resulting ANN system can accurately classify four ECG types: normal, atrial premature contraction, left bundle branch block, and premature ventricular contraction. The performance of the ECG classifier models is evaluated based on accuracy and power. Among the three AI algorithms, the SNN requires the lowest power consumption of 0.226 W on-chip, followed by MLP (1.677 W), and CNN (2.266 W). However, the highest accuracy is achieved by the CNN (95%), followed by MLP (76%) and SNN (90%).

• Current Photovoltaic Research
• /
• v.12 no.3
• /
• pp.49-54
• /
• 2024

This study proposed a forecasting model that combines ANNs and RNNs to address the intermittency and fluidity of solar power generation. Four prediction models were trained separately based on sky conditions provided by the Korea Meteorological Administration, and insolation was estimated using the ASHRAE Clear-Sky model. The proposed model showed an error rate of 6.5-7.7% based on NMAE, which meets the requirements of power generation prediction. As a result, this study can improve the accuracy of solar power generation forecasting, which can contribute to the stability of power operation and the profitability of power operators.

• Steel and Composite Structures
• /
• v.17 no.5
• /
• pp.623-639
• /
• 2014

Due to recent advancements in the area of Artificial Intelligence (AI) and computational intelligence, the application of these technologies in the construction industry and structural analysis has been made feasible. With the use of the Adaptive-Network-based Fuzzy Inference System (ANFIS) as a modelling tool, this study aims at predicting the shear strength of channel shear connectors in steel concrete composite beam. A total of 1200 experimental data was collected, with the input data being achieved based on the results of the push-out test and the output data being the corresponding shear strength which were recorded at all loading stages. The results derived from the use of ANFIS and the classical linear regressions (LR) were then compared. The outcome shows that the use of ANFIS produces highly accurate, precise and satisfactory results as opposed to the LR.