Earticle

1,2인 가구가 꾸준하게 늘어나고 있는 추세에 비어 있는 시간대에 집을 방문하는 외부인이 누구인지 확인하고 싶은 요구가 증가하고 있다. 얼굴인식 기술은 많은 연구를 통해 여러 가지 모델이 제안되었는데 OpenCV의 Harr Cascade와 Dlib의 Hog가 대표적인 오픈소스 모델이다. 두 모델은 사용 환경에 따른 장단점을 가지고 있는데, 본 연구 에서 초점을 둔 실내 현관 앞과 제한된 거리에서는 Dlib의 Hog가 강점을 가진다. 본 논문에서는 딥러닝 오픈 소스인 Dlib에 기반을 둔 얼굴인식 방문자 출입관리 시스템을 설계하고 구현하였다. 전체 시스템은 프론트 모듈과 서버모듈, 모바일모듈로 구성되며 세부적으로는 얼굴등록, 얼굴인식, 실시간 방문자 확인 및 원격제어, 동영상 저장 기능을 포함한 다. 인터넷에서 공개된 사진을 이용하여 거리임계 값의 변화에 따른 정밀도, 특이도, 정확도를 구하고 선행연구 결과와 비교하였다. 실험 결과 구현된 시스템이 정상적으로 동작하는 것을 확인 하였으며 Dlib에서 보고한 것과 비슷한 결과를 보이는 것을 확인 하였다.

As the trend of steadily increasing the number of single or double household, there is a growing demand to see who is the outsider visiting the home during the free time. Various models of face recognition technology have been proposed through many studies, and Harr Cascade of OpenCV and Hog of Dlib are representative open source models. Among the two modes, Dlib's Hog has strengths in front of the indoor and at a limited distance, which is the focus of this study. In this paper, a face recognition visitor access system based on Dlib was designed and implemented. The whole system consists of a front module, a server module, and a mobile module, and in detail, it includes face registration, face recognition, real-time visitor verification and remote control, and video storage functions. The Precision, Specificity, and Accuracy according to the change of the distance threshold value were calculated using the error matrix with the photos published on the Internet, and compared with the results of previous studies. As a result of the experiment, it was confirmed that the implemented system was operating normally, and the result was confirmed to be similar to that reported by Dlib.

This paper presents an artificial intelligence-based algorithmic trading simulation platform designed for the Vietnamese market. First, we gather and merge a wide range of financial information, including stock market indicators, gold price movements, corporate earnings reports, and financial news. Apache Airflow facilitates the scheduling and automation of the extract, transform, load (ETL) tasks for data collection and transport. Using the processed dataset, the platform employs several learning models including random forests and long short-term memory networks to generate short-term forecasts. The predictive outputs are then combined with rule-based mechanisms such as momentum rules and pairs-trading logic to produce trading signals. In addition to making predictions, the system acts as a practical educational platform. This enables students and researchers to design, test, and compare trading strategies. The system reports detailed performance statistics, including risk-adjusted returns, maximum drawdown, portfolio turnover, and hit ratios, thereby supporting transparent performance evaluation.

With the growth of Internet-of-Things (IoT) technologies, the number of IoT devices developers need to manage has increased exponentially. Many IoT standards have been proposed to allow those devices to communicate efficiently in day-to-day tasks. However, we lack trusted interworking entities for devices from different standards to collaborate securely. This paper proposes a blockchain platform that bridges multiple heterogeneous IoT platforms to co-exist and interwork. Specifically, we designed an interworking proxy application entity (IPE) implemented as a chaincode in Hyperledger Fabric to collect and process data coming from/to oneM2M and LWM2M architecture. The use of blockchain will guarantee network reliability and data integrity so that cross-standard communications can be audited and processed securely. Based on our evaluation, we show that the interworking between oneM2M and LWM2M through our blockchain platform is feasible. Furthermore, the proposed system can process up to 206 transactions per second with 1,000 running applications, which is about an 87% increase from the previously referenced study.

This paper presents a novel method to design and implement mobile u-health system by defining the essential elements of mobile healthcare services. We choose common service elements for the proposed u-healthcare scheme and design the service platform. Especially we focus on automatic feces or urine sensing u-care scheme to prove the effectiveness of our platform. We construct the system with sensing part with a manikin and a diaper wireless communication part with feces or urine sensing data, and coordinator system based on the u-health platform defined in this paper. Experimental results show that our scheme is useful in the area of u-care service for the handicapped, the elderly, and patients who can hardly move by themselves. In addition the designed scheme offers a realized u-care scheme with the purpose of advanced developing tools for application or service developers.

Recently, RFID technology has attracted considerable attention in many industry fields. The RFID environment requires a standard architecture for the smooth exchange of data between heterogeneous networks. The architecture should offer an efficient standard environment, such as a communication environment based on Web Services, PKI or Kerberos-based security, and abstract business processes which could be used in the diverse domains. Therefore, in this paper, we propose an Enterprise Application Framework (EAF) which includes a standard communication protocol, security functions, and abstract level business processes. The suggested architecture is expected to provide a more secure and flexible security management in the dynamic RFID application environments, and is expected to provide an abstract business event for the development of business processes which could apply RFID technology to the existing systems.

This study presents an adaptive English text regeneration system that modifies authentic materials to match learners' CEFR (Common European Framework of Reference for Languages) proficiency levels (A1-C2). It addresses the crucial challenge of accessibility while maintaining the original meaning. Leveraging advancements in large language models (LLMs), our framework employs a three-phase process: first, CEFR-based text analysis utilizing curated vocabulary lists and syntactic metrics; second, multi-level regeneration through the fine-tuned Qwen 2.5 model; and third, rigorous validation of semantic fidelity (achieving a 92% BERT score) and readability. Experimental results with 300 learners indicate significant improvements, with a 32% increase in comprehension for beginner groups and a 25% increase for intermediate groups. Additionally, there is a 40% decrease in self-reported anxiety. The system's real-time processing capability (under 3 seconds per page) ensures practical scalability.Our work makes three key contributions: it establishes the first comprehensive framework covering all six CEFR levels with empirical validation; it integrates pedagogical and psychological principles to boost learner motivation and reduce anxiety; and it demonstrates the effectiveness of progressive complexity scaffolding while setting actionable benchmarks for LLM-driven educational tools. By balancing linguistic precision with psychological benefits-such as increased motivation and confidence-the system enhances the role of AI in language education. Future research will focus on adapting colloquial language and examining longitudinal impacts on knowledge retention, further bridging the gap between authentic content and learner needs.

An electronic document based on PDF has been widely used in customer communication between an enterprise and a customer to deliver personalized content. However, electronic documents based on PDF in the form of paper layouts are not suitable for mobile environments because of low readability and lack of interactive interaction. Even though HTML is an essential language in a mobile environment, electronic document based on PDF is still used as it has a content integrity verification feature with a digital signature. It means that a user is sacrificing user experience in a mobile environment for content integrity and using paper-layout electronic documents. In this research, we design the Document HTML specification by setting the Document HTML conformance, adding the extended meta tags, and signing the message digest with a digital signature based on public key infrastructure (PKI). Furthermore, we implemented the Document HTML system, which has REST API services to generate and verify the Document HTML, and did experimental verification of the theory. As a result, we have confirmed that the Document HTML has both content integrity and user experience on mobile. Furthermore, the Document HTML is expected to be an alternative document format to deliver personalized content from an enterprise to a customer in a mobile environment instead of the paper layout electronic document such as PDF.

Recently, as various examples of machine learning have been applied in the healthcare field, deep learning technology has been applied to various tasks, such as electrocardiogram examination and body composition analysis using wearable devices such as smart watches. To utilize deep learning, securing data is the most important procedure, where human intervention, such as data classification, is required. In this study, we propose a model that uses a clustering algorithm, namely, the K-means clustering, to label body fat according to gender and age considering body size aspects, such as chest circumference and waist circumference, and classifies body fat into five groups from high risk to low risk using a convolutional neural network (CNN). As a result of model validation, accuracy, precision, and recall results of more than 95% were obtained. Thus, rational decision making can be made in the field of healthcare or obesity analysis using the proposed method.

In this paper, to prevent accidents of school vehicles used by students, which have emerged as a continuous social problem, a boarding and leaving notification system using Bluetooth beacons is proposed. The platform structure uses Bluetooth, a smartphone, and an application, and provides real-time message reception between parents and teachers. It is not simply a system with Bluetooth capability alone, but also an IcT technology that is closely related to real life using two-way communication. The purpose is to provide user-friendly ICT technology as an integrated system for more practical effects. Two-way exchange of information between parents and teachers using cloud servers is made possible. By using this Bluetooth beacon notification system, safety accidents in student vehicles are reduced, which brings confidence from traffic accidents to schoolchildren of parents and teachers. Therefore, it is expected that parents and teachers are expected to improve their mental stability and quality of life from traffic accidents.

When a crime occurs, the information necessary for solving the case, and various pieces of the evidence needed to prove the crime are collected from the crime scene. The tangible residues collected through scientific methods at the crime scene become evidence at trial and a clue to prove the facts directly against the offense of the suspect. Therefore, the scientific investigation and forensic handling for securing objective forensic in crime investigation is increasingly important. Today, digital systems, such as smartphones, CCTVs, black boxes, etc. are increasingly used as criminal information investigation clues, and digital forensic is becoming a decisive factor in investigation and trial. However, the systems have the risk that digital forensic may be damaged or manipulated by malicious insiders in the existing centralized management systems based on client/server structure. In this paper, we design and implement a blockchain based digital forensic management model using Hyperledger Fabric and Docker to guarantee the reliability and integrity of digital forensic. The proposed digital evidence management model allows only authorized participants in a distributed environment without a central management agency access the network to share and manage potential crime data. Therefore, it could be relatively safe from malicious internal attackers compared to the existing client/server model.

We usually suffer from difficulties in treating or managing Big Data generated from various digital media and/or sensors using traditional mining techniques. Additionally, there are many problems relative to the lack of memory and the burden of the learning curve, etc. in an increasing capacity of large volumes of text when new data are continuously accumulated because we ineffectively analyze total data including data previously analyzed and collected. In this paper, we propose a general-purpose classifier and its structure to solve these problems. We depart from the current feature-reduction methods and introduce a new scheme that only adopts changed elements when new features are partially accumulated in this free-style learning environment. The incremental learning module built from a gradually progressive formation learns only changed parts of data without any re-processing of current accumulations while traditional methods re-learn total data for every adding or changing of data. Additionally, users can freely merge new data with previous data throughout the resource management procedure whenever re-learning is needed. At the end of this paper, we confirm a good performance of this method in data processing based on the Big Data environment throughout an analysis because of its learning efficiency. Also, comparing this algorithm with those of NB and SVM, we can achieve an accuracy of approximately 95% in all three models. We expect that our method will be a viable substitute for high performance and accuracy relative to large computing systems for Big Data analysis using a PC cluster environment.

For 3D display applications, auto-stereoscopic display methods that can provide 3D images without glasses have been actively developed. This paper is concerned with developing a display system for elemental images of real space using integral imaging. Unlike the conventional method, which reduces a color image to the level as much as a generated depth image does, we have minimized original color image data loss by generating an enlarged depth image with interpolation methods. Our method was efficiently implemented by applying a GPU parallel processing technique with OpenCL to rapidly generate a large amount of elemental image data. We also obtained experimental results for displaying higher quality integral imaging rather than one generated by previous methods.

A wireless sensor network (WSN) is designed and implemented for a freeze dryer. Freeze-drying technology is widely used in the fields of pharmacy and biotechnology as well as the food and agriculture industries. Taking into account the demand for high-resolution pressure and temperature measurements in a freeze dryer, the proposed WSN has a significant advantage of creating a monitoring environment in a freeze dryer. The proposed WSN uses a ZigBee/IEEE 802.15.4 network with an altimeter module that contains a high-resolution pressure and temperature sensor with a serial digital data interface. The ZigBee network is suitable for low-energy and low-data-rate applications in the field of wireless communication. The altimeter module is capable of sensing pressure in the range of 7.5-975 Torr (10-1300 mbar) and temperature in the range of $-40^{\circ}C$ to $125^{\circ}C$ with a DC power consumption of $3{\mu}W$. The implemented WSN is installed in a commercial laboratory freeze dryer in order to demonstrate its functionality and efficiency. A comparison with the temperature profile measured by a thermocouple installed in the freeze dryer reveals that the resolution of the temperature profile measured by WSN is superior to that measured by the thermocouple.

This paper proposes a method to manipulate digital hologram contents by manipulating and/or synthesizing the depth information. To synthesize digital holograms themselves in order to create new digital hologram contents. This paper uses both the depth information obtained by converting the disparity information by using a stereo matching method and that obtained by taking pictures with a depth camera. In addition, assuming that digital holograms are created using the computer-generated holography method, we propose a technique for authoring and compositing hologram contents by using either the changes in the three-dimensional positions of objects in the hologram or by combining the objects with other contents by means of changes in the depth information. Further, more than one digital hologram was synthesized to form a hologram. The reconstructed result from the synthesized hologram also contained all the objects in each digital hologram before synthesis at the same positions and distances.

It is difficult for mobile learners to maintain a high level of concentration when learning content for more than an hour while they are on the move. Despite the attention span issue, many m-learning systems still provide their mobile learners with the same content once used in e-learning systems. This has called for an investigation to identify the suitable characteristics of the m-learning environment. With this in mind, we have conducted a survey in hopes of determining the requirements for developing more suitable m-learning content. Based on the results of the survey, we have developed a content model comprised of two types: a segment type and a supplement type. In addition, we have implemented a prototype system of the content model for Apple iPhones and Android smartphones in order to investigate a feasibility study of the model application.

An intelligent light emitting diode (LED) dimming system is designed and implemented for energy-saving lighting systems. An LED light bulb is powered by an LED driver controlled by a microcontroller using pulse width modulation (PWM) signals. By changing the duty cycle of the PWM signals, the LED driver generates a driving current of up to 1,000 mA. The current consumption by the LED light bulb exhibits a very linear characteristic that indicates that the level of LED dimming can be finely tuned. Multiple sensors-lighting intensity and ultrasonic range sensors-are combined with the LED dimming system to realize an automatically controllable LED lighting system. The light intensity sensor is capable of sensing ambient light. The ultrasonic range sensor can detect objects from 0.15 to 5.6 m at a resolution of 0.0254 m. The collected information by the light intensity and ultrasonic range sensors is processed by the microcontroller that in turn automatically controls the brightness of the LED light bulb. The algorithm of the software for the LED dimming system is also described.

RFID technology is one of the verification technology that use radio frequency. Use and study of RFID are on the increase because RFID is a one of main technologies in ubiquitous. In this paper, we address to develop a Record Processing System using RFID. The purpose of this paper is improvement of accuracy and efficiency on Record Processing System by applying RFID. Later, this paper will be able to help to establish of u-Campus.

Recently, biological researches are required to deal with a large scale of data. While scientists used classical experimental approaches for researches in the past, it is possible to get more sophisticated observations easily with the convergence of information technologies and biology. The study on diseases is one of the most important issues of the life science. Conventional services and databases provide users with information such as classification of diseases, symptoms, and medical treatments through the Web. However, it is hard to connect or develop them for other new services because they have independent and different criteria. It may be a factor that interferes the development of biology. In this paper, we propose integrated data structures for the disease databases. We also design and implement a novel directory system for diseases as an infrastructure for developing the new diseases services.

The main memory DBMS (MMDBMS) efficiently supports various database applications that require high performance since it employs main memory rather than disk as a primary storage. An index manager is an essential sub-component of a DBMS used to speed up the retrieval of objects from a large volume of a database in response to a certain search condition. Previous research efforts on indexing proposed various index structures. However, they hardly dealt with the practical issues occurred in implementing an index manager on a target DBMS. In this paper, we touch these issues and present our experiences in developing the index manager. The main issues are (1) compact representation of an index entry, (2) support of variable-length keys. (3) support of multiple-attribute keys, and (4) support of duplicated keys.

Robot has been developed as something to help human beings, and robot will be very important field as an entertainment. This paper intends to contribute to developing golf robot as an entertainment using AVR ATmega Chips. It is believed that the entertaining robot giving a pleasure to human beings has good prospects for the rapid growth along with other robot industry. Golf robot developed in this paper has been developed for entertainment and has a similar sense in comparison with the real golf. This golf robot is represented in the way of putting the ball in the hall by hitting the ball just like the actual golf game and putting through making robot come close to the ball with the putting robot. This golf robot can playa game with several people like the actual golf if many putting robots can be used. It is, therefore, considered that the development of golf robots for entertainment using AVR has a high value of the golf robot as an entertainer on the ground that it can playa real golf.