• Nuclear Engineering and Technology
• /
• v.55 no.2
• /
• pp.443-451
• /
• 2023

Machine Learning has introduced many solutions in data science, but its application in IoT faces significant challenges, due to the limitations in memory size and processing capability of constrained devices. In this paper we design an automatic gamma radiation detection and identification embedded system that exploits the power of TinyML in a SiPM micro radiation sensor leveraging the Edge Impulse platform. The model is trained using real gamma source data enhanced by software augmentation algorithms. Tests show high accuracy in real time processing. This design has promising applications in general-purpose radiation detection and identification, nuclear safety, medical diagnosis and it is also amenable for deployment in small satellites.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.26 no.2
• /
• pp.93-98
• /
• 2016

For accurate precipitation forecasts the choice of weather factors and prediction method is very important. Recently, machine learning has been widely used for forecasting precipitation, and artificial neural network, one of machine learning techniques, showed good performance. In this paper, we suggest a new method for forecasting precipitation using DBN, one of deep learning techniques. DBN has an advantage that initial weights are set by unsupervised learning, so this compensates for the defects of artificial neural networks. We used past precipitation, temperature, and the parameters of the sun and moon's motion as features for forecasting precipitation. The dataset consists of observation data which had been measured for 40 years from AWS in Seoul. Experiments were based on 8-fold cross validation. As a result of estimation, we got probabilities of test dataset, so threshold was used for the decision of precipitation. CSI and Bias were used for indicating the precision of precipitation. Our experimental results showed that DBN performed better than MLP.

• Journal of IKEEE
• /
• v.20 no.4
• /
• pp.398-411
• /
• 2016

SVM, a machine learning method, is widely used in image processing for it's excellent generalization performance. However, to add other data to the pre-trained data of the system, we need to train the entire system again. This procedure takes a lot of time, especially in embedded environment, and results in low performance of SVM. In this paper, we implemented an SVM trainer and classifier in an FPGA to solve this problem. We parlallelized the repeated operations inside SVM and modified the exponential operations of the kernel function to perform fixed point modelling. We implemented the proposed hardware on Xilinx ZC 706 evaluation board and used TSR algorithm to verify the FPGA result. It takes about 5 seconds for the proposed hardware to train 2,000 data samples and 16.54ms for classification for $1360{\times}800$ resolution in 100MHz frequency, respectively.

• Journal of the Korea Society of Computer and Information
• /
• v.17 no.3
• /
• pp.77-83
• /
• 2012

As web pages provide various services, the distribution of malware via the web pages is being also increased. Malware can make personal information leak, system mal-function and system be zombie. To protect this damages, we should block the malicious web pages. Because the malicious codes embedded in web pages are obfuscated or transformed, it is difficult to detect them using signature-based approaches which are used by current anti-virus software. To overcome this problem, we extracted features to classify malicious web pages and benign ones by analyzing web pages. And we propose a classification method using SVM which is widely used in machine learning. Experimental results show that the proposed method is better than other methods. The proposed method could classify malicious web pages correctly and be helpful to block the distribution of malicious codes.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.26 no.5
• /
• pp.668-674
• /
• 2022

Machine learning techniques using visual data have high usability in fields of industry and service such as scene recognition, fault detection, security and user analysis. Among these, user analysis through the videos from CCTV is one of the practical way of using vision data. Also, many studies about lightweight artificial neural network have been published to increase high usability for mobile and embedded environment so far. In this study, we propose the network combining the object detection and classification for mobile graphic processing unit. This network detects pedestrian and face, classifies age and gender from detected face. Proposed network is constructed based on MobileNet, YOLOv2 and skip connection. Both detection and classification models are trained individually and combined as 2-stage structure. Also, attention mechanism is used to improve detection and classification ability. Nvidia Jetson Nano is used to run and evaluate the proposed system.

• Journal of the Korea Society of Computer and Information
• /
• v.29 no.3
• /
• pp.75-81
• /
• 2024

In this paper, we proposes a method for real-time processing of inter-floor noise problems by embedding TinyML, which includes a deep learning model, into ultra-low-power systems. The reason this method is feasible is because of lightweight deep learning model technology, which allows even systems with small computing resources to perform inference autonomously. The conventional method proposed to solve inter-floor noise problems was to send data collected from sensors to a server for analysis and processing. However, this centralized processing method has issues with high costs, complexity, and difficulty in real-time processing. In this paper, we address these limitations by employing On-Sensor AI using TinyML. The method presented in this paper is simple to install, cost-effective, and capable of processing problems in real-time.

• Journal of Intelligence and Information Systems
• /
• v.29 no.3
• /
• pp.419-437
• /
• 2023

As the number of thin filers in Korea surpasses 12 million, there is a growing interest in enhancing the accuracy of assessing their credit default risk to generate additional revenue. Specifically, researchers are actively pursuing the development of default prediction models using machine learning and deep learning algorithms, in contrast to traditional statistical default prediction methods, which struggle to capture nonlinearity. Among these efforts, Graph Neural Network (GNN) architecture is noteworthy for predicting default in situations with limited data on thin filers. This is due to their ability to incorporate network information between borrowers alongside conventional credit-related data. However, prior research employing graph neural networks has faced limitations in effectively handling diverse categorical variables present in credit information. In this study, we introduce the Transformer embedded Graph Convolutional Network (TeGCN), which aims to address these limitations and enable effective default prediction for thin filers. TeGCN combines the TabTransformer, capable of extracting contextual information from categorical variables, with the Graph Convolutional Network, which captures network information between borrowers. Our TeGCN model surpasses the baseline model's performance across both the general borrower dataset and the thin filer dataset. Specially, our model performs outstanding results in thin filer default prediction. This study achieves high default prediction accuracy by a model structure tailored to characteristics of credit information containing numerous categorical variables, especially in the context of thin filers with limited data. Our study can contribute to resolving the financial exclusion issues faced by thin filers and facilitate additional revenue within the financial industry.

• Journal of IKEEE
• /
• v.20 no.4
• /
• pp.351-360
• /
• 2016

Recently, machine learning algorithms, especially deep learning-based algorithms, have been receiving attention due to its high classification performance. Among the algorithms, Convolutional Neural Network(CNN) is known to be efficient for image processing tasks used for Advanced Driver Assistance Systems(ADAS). However, it is difficult to achieve real-time processing for CNN in vehicle embedded software environment due to the repeated operations contained in each layer of CNN. In this paper, we propose a hardware accelerator which enhances the execution time of CNN by parallelizing the repeated operations such as convolution. Xilinx ZC706 evaluation board is used to verify the performance of the proposed accelerator. For $36{\times}36$ input images, the hardware execution time of CNN is 2.812ms in 100MHz clock frequency and shows that our hardware can be executed in real-time.

• Journal of Biomedical Engineering Research
• /
• v.37 no.5
• /
• pp.161-167
• /
• 2016

Recent technological advances in sensor fabrication and bio-signal processing enabled non-constraint and non-intrusive measurement of human bio-signals. Especially, non-constraint measurement of ECG makes it available to estimate various human health parameters such as heart rate. Additionally, non-constraint ECG measurement of wheelchair user provides real-time health parameter information for emergency response. For accurate emergency response with low false alarm rate, it is necessary to discriminate quality levels of ECG measured using non-constraint approach. Health parameters acquired from low quality ECG results in inaccurate information. Thus, in this study, a machine learning based approach for three-class classification of ECG quality level is suggested. Three sensors are embedded in the back seat, chest belt, and handle of automatic wheelchair. For the two sensors embedded in back seat and chest belt, capacitively coupled electrodes were used. The accuracy of quality level classification was estimated using Monte Carlo cross validation. The proposed approach demonstrated accuracy of 94.01%, 95.57%, and 96.94% for each channel of three sensors. Furthermore, the implemented algorithm enables classification of user posture by detection of contacted electrodes. The accuracy for posture estimation was 94.57%. The proposed algorithm will contribute to non-constraint and robust estimation of health parameter of wheelchair users.

• Smart Media Journal
• /
• v.10 no.3
• /
• pp.31-38
• /
• 2021

It is hard to predict when and where a fall accident will happen. Also, if rapid follow-up measures on it are not performed, a fall accident leads to a threat of life, so studies that can automatically detect a fall accident have become necessary. Among automatic fall-accident detection techniques, a fall detection scheme using an IMU (inertial measurement unit) sensor attached to a wrist is difficult to detect a fall accident due to its movement, but it is recognized as a technique that is easy to wear and has excellent accessibility. To overcome the difficulty in obtaining fall data, this study proposes an algorithm that efficiently learns less data through machine learning such as KNN (k-nearest neighbors) and SVM (support vector machine). In addition, to improve the performance of these mathematical classifiers, this study utilized feature data aquired in the frequency space. The proposed algorithm analyzed the effect by diversifying the parameters of the model and the parameters of the frequency feature extractor through experiments using standard datasets. The proposed algorithm could adequately cope with a realistic problem that fall data are difficult to obtain. Because it is lighter than other classifiers, this algorithm was also easy to implement in small embedded systems where SIMD (single instruction multiple data) processing devices were difficult to mount.